JGV Papers in Press. Published May 29, 2015 as doi:10.1099/vir.0.000203
Journal of General Virology The role of cytomegalovirus (CMV)-specific polyfunctional CD8+ T cells and antibodies neutralizing virus epithelial infection in the control of CMV infection in the allogeneic stem cell transplantation setting. --Manuscript Draft-Manuscript Number:
VIR-D-15-00162R2
Full Title:
The role of cytomegalovirus (CMV)-specific polyfunctional CD8+ T cells and antibodies neutralizing virus epithelial infection in the control of CMV infection in the allogeneic stem cell transplantation setting.
Short Title:
CMV-specific neutralizing antibodies and CMV DNAemia
Article Type:
Standard
Section/Category:
Animal - Large DNA Viruses
Corresponding Author:
David Navarro School of Medicine, University of Valencia, Spain Valencia, SPAIN
First Author:
Estela Giménez
Order of Authors:
Estela Giménez Pilar Blanco-Lobo Beatriz Muñoz-Cobo Carlos Solano Paula Amat Pilar Pérez-Romero David Navarro
Abstract:
The role cytomegalovirus (CMV)-specific polyfunctional CD8+ T cells and that of antibodies neutralizing virus epithelial infection (AbNEI) in the control of CMV DNAemia was investigated in 39 CMV-seropositive allogeneic stem cell transplant recipients (Allo-SCT) with (n=24) or without (n=15) CMV DNAemia. AbNEI levels were prospectively monitored by means of a neutralization assay employing retinal epithelial cells (ARPE-19) and the recombinant CMV strain BADrUL131-Y4. Quantification of CMV-specific polyfunctional CD8+ T cells (expressing 2 or 3 of the following markers: IFN-γ, TNF-α, CD107a) in whole blood was performed by flow cytometry for intracellular cytokine staining. We found no differences in the dynamic pattern of AbNEI in patients with or without subsequent CMV DNAemia. Baseline and peak AbNEI titers were not predictive of the dynamics of CMV replication within episodes. No correlation was found between CMV DNA loads and AbNEI levels during episodes of CMV DNAemia (rho=0.09; 95% C.I. -0.52-0.64; P=0.78). The detection of pp65/IE-1 CMVspecific polyfunctional CD8+ T cells was associated with low level viral replication within subsequent episodes of CMV DNAemia. Interestingly, the presence of AbNEI titers (inverse) >4.7 log2 was predictive of the occurrence of CMV DNAemia (sensitivity, 83%; Specificity, 80%). Our findings should add insights to the role of humoral and cellular immunity in the control of CMV infection in the Allo-SCT setting.
Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Manuscript Including References (Word document) Click here to download Manuscript Including References (Word document): PP DN JGV 2015 text tables revised.2 definitive.doc
1
VIR-D-15-00162.R2
2
The role of cytomegalovirus (CMV)-specific polyfunctional CD8+ T cells and
3
antibodies neutralizing virus epithelial infection in the control of CMV infection in
4
the allogeneic stem cell transplantation setting
5
Estela Giménez,1# Pilar Blanco-Lobo,2# Beatriz Muñoz-Cobo,1 Carlos Solano,3 Paula
6
Amat,3 Pilar Pérez-Romero,2 and David Navarro1,4
7
1
8
Spain.
9
2
Microbiology Service, Hospital Clínico Universitario, Fundación INCLIVA, Valencia,
Instituto
de
Biomedicina
de
Sevilla,
Hospital
Universitario
Virgen
del
10
Rocío/CSIC/Universidad de Sevilla, Sevilla, Spain.
11
3
12
INCLIVA, Valencia, Spain.
13
4
14
Spain.
15
#Authors contributed equally to this work.
16
Author for correspondence: David Navarro. Microbiology Service, Hospital Clínico
17
Universitario, and Department of Microbiology, School of Medicine, Valencia, Spain.
18
Av. Blasco Ibáñez 17, 46010 Valencia, Spain. Phone: 34(96)3864657; Fax:
19
34(96)3864173; E-mail: [email protected]
20
Running Title: CMV-specific neutralizing antibodies and CMV DNAemia.
Hematology and Medical Oncology Service, Hospital Clínico Universitario, Fundación
Department of Microbiology, School of Medicine, University of Valencia, Valencia,
1
21
Word counts. Summary: 216; Text: 3,541.
22
SUMMARY
23
The role cytomegalovirus (CMV)-specific polyfunctional CD8+ T cells and that of
24
antibodies neutralizing virus epithelial infection (AbNEI) in the control of CMV
25
DNAemia was investigated in 39 CMV-seropositive allogeneic stem cell transplant
26
recipients (Allo-SCT) with (n=24) or without (n=15) CMV DNAemia. AbNEI levels
27
were prospectively monitored by means of a neutralization assay employing retinal
28
epithelial cells (ARPE-19) and the recombinant CMV strain BADrUL131-Y4.
29
Quantification of CMV-specific polyfunctional CD8+ T cells (expressing 2 or 3 of the
30
following markers: IFN-γ, TNF-α, CD107a) in whole blood was performed by flow
31
cytometry for intracellular cytokine staining. We found no differences in the dynamic
32
pattern of AbNEI in patients with or without subsequent CMV DNAemia. Baseline and
33
peak AbNEI titers were not predictive of the dynamics of CMV replication within
34
episodes. No correlation was found between CMV DNA loads and AbNEI levels during
35
episodes of CMV DNAemia (rho=0.09; 95% C.I. -0.52-0.64; P=0.78). The detection of
36
pp65/IE-1 CMV-specific polyfunctional CD8+ T cells was associated with low level
37
viral replication within subsequent episodes of CMV DNAemia. Interestingly, the
38
presence of AbNEI titers (inverse) >4.7 log2 was predictive of the occurrence of CMV
39
DNAemia (sensitivity, 83%; Specificity, 80%). Our findings should add insights to the
40
role of humoral and cellular immunity in the control of CMV infection in the Allo-SCT
41
setting.
2
42
Key Words: Cytomegalovirus (CMV), CMV DNAemia, Allogeneic stem cell
43
transplantation (Allo-SCT), neutralizing antibodies, CMV-specific polyfunctional CD8+
44
T cells.
45
INTRODUCTION
46
Protection from and clearance of cytomegalovirus (CMV) viremia in allogeneic stem
47
cell transplant recipients (Allo-SCT) is critically dependent upon the reconstitution and
48
adequate expansion, respectively, of functional CMV-specific T cells (Solano &
49
Navarro, 2010). Whether CMV-specific antibodies exerting antiviral activities
50
contribute in a relevant way to these purposes has not been elucidated. In this context,
51
the
52
hyperimmunoglobulin on minimizing the incidence and fatal outcome of CMV end-
53
organ disease, claimed in this clinical setting, has never been conclusively proven
54
(Guglielmo et al., 1994). Furthermore, contradictory data have been published on the
55
role of CMV-specific neutralizing antibodies in the prevention or control of CMV
56
viremia (Muñoz et al., 2001; Schoppel et al., 1998; Volpi et al., 1999); In these studies,
57
neutralizing antibodies were quantified by means of a classical neutralization assay
58
involving the use of the CMV AD169 strain and human-derived fibroblasts, which
59
mostly measures antibody levels against gB and the glycoprotein complex gH–gL, both
60
essential for infection of human fibroblasts (Vanarsdall et al., 2008). Recently, the
61
virion glycoprotein complex gH–gL–pUL131A–pUL130–pUL128 has been shown to
62
be required for infection of endothelial, epithelial and myeloid cells (Gerna et al., 2005;
63
Hahn et al., 2004; Ryckman et al., 2008a,b; Wang & Shenk, 2005a), all of which are
64
key targets for CMV pathogenesis in the human host. Neutralizing antibodies targeting 3
beneficial
effect
of
the
administration
of
CMV-specific
human
65
this pentameric complex are known to display a much higher potency in neutralizing
66
virus infection of epithelial and endothelial cells than those against gB (Macagno et al.,
67
2010). The AD169 strain of CMV carries a defective UL131 ORF, thus antibodies
68
against this pentameric complex are not measured in the classical neutralization assay
69
(Hahn et al., 2004). Antibodies against neutralizing epitopes of the complex gH–gL–
70
pUL131A–pUL130–pUL128 have been shown to correlate with virus control in
71
immunocompetent individuals (Lilleri et al., 2012) and protection from fetal CMV
72
transmission (Lilleri et al., 2013). Here, we investigated in parallel the role of CMV-
73
specific antibodies that neutralize epithelial cell infection (AbNEI) and that of
74
cytomegalovirus (CMV)-specific polyfunctional CD8+ T in the control of CMV
75
DNAemia in a cohort of Allo-SCT recipients.
76
RESULTS
77
Incidence of CMV DNAemia. A total of 24 out of 39 patients (61.5%) developed a
78
first episode of CMV DNAemia within the first 100 days after Allo-SCT (median 30
79
days; Range 26-42 days). All episodes were eventually controlled. Twelve episodes
80
required the administration of pre-emptive antiviral therapy. CMV DNAemia was
81
spontaneously cleared in the remaining 12 patients.
82
Kinetics of CMV-specific antibodies neutralizing epithelial infection (AbNEI) in
83
patients with or without CMV DNAemia. We first assessed the kinetics of AbNEI in
84
patients with or without subsequent CMV DNAemia. Sequential sera (n=60) were
85
available from 18 patients subsequently having CMV DNAemia (median, 3/patient;
86
Range, 2-8/patient). Serial sera (n=76) from 15 patients with no CMV DNAemia
87
(median, 5/patient; Range, 3-8/patient) obtained within a comparable time frame were 4
88
also available for analysis and served as controls. The kinetic profiles of AbNEI did not
89
differ notably between patients in the study groups (Figure 1, panels A and B). CMV-
90
specific AbNEI levels were found to remain stable in 13/18 and 10/15 patients with or
91
without subsequent CMV DNAemia, respectively (P=0.73). Either increasing,
92
decreasing or fluctuating levels of AbNEI levels were observed at comparable
93
frequencies among individuals belonging to both groups (P=>0.5). Data on CMV CMV
94
(AD169)-specific IgG levels, as measured by CLIA, were available from 31 patients (a
95
median of 3 samples/patient; Range, 1-6). All sera having detectable AbNEI were also
96
reactive by the CMV IgG CLIA. As shown in Figure 1 (panels C and D), the kinetics
97
pattern of CMV IgG antibodies was found to be comparable in patients with or without
98
subsequent CMV DNAemia, with a rather similar number of patients displaying stable,
99
increasing or decreasing antibody levels in both study groups.
100
Quantitation of CMV-specific antibodies neutralizing epithelial infection allows for
101
anticipation of the occurrence of CMV DNAemia. We next investigated whether
102
AbNEI levels measured either at the time of or early after Allo-SCT were indicative of
103
the risk of CMV DNAemia. The data for each patient of the cohort are shown in Table
104
1.We found that both baseline (median, day 0; Range, day -1 to 16) and peak (median,
105
day 14; Range, days 0-32) AbNEI levels were significantly higher in patients who went
106
on to develop CMV DNAemia than in those who did not (Figure 2, panels A and B,
107
respectively). ROC curve analyses indicated that a cut-off of 4.7 log2 titer (1/120) for
108
both baseline and peak AbNEI discriminated fairly well (sensitivity, 83%; specificity,
109
80%) between patients with or without subsequent CMV DNAemia (Figure 3). In fact,
110
the odds ratio for developing CMV DNAemia was 20.0 (95% C.I 3.81-105.11) for
111
patients with AbNEI titers above the referred threshold. In contrast, both baseline and 5
112
peak CMV-specific IgG titers were comparable in both study groups (Figure 2, panels C
113
and D). Thus, a CMV IgG threshold level discriminating between patients with or
114
without subsequent CMV DNAemia could not be established. Of note the fact that
115
patients displaying baseline and peak AbNEI titers either above or below this cut-off
116
were comparable in terms of pre-transplant and post-transplant clinical factors known to
117
modulate the risk of CMV DNAemia (not shown).
118
The dynamics of CMV DNAemia is not influenced by CMV-specific antibodies
119
neutralizing epithelial infection. We then investigated whether AbNEI levels
120
measured prior to the occurrence of CMV DNAemia allowed for the inference of the
121
dynamics of virus replication within episodes. As shown in Table 2, neither initial nor
122
peak plasma CMV DNA loads differed significantly between patients displaying
123
baseline or peak AbNEI levels above or below 4.7 log2 titer. Likewise, both the need for
124
pre-emptive antiviral therapy and the duration of episodes of CMV DNAemia were not
125
associated with AbNEI levels. Furthermore, no significant correlation was observed
126
between peak AbNEI levels and peak plasma CMV DNA loads (rho=-0.27; 95% C.I., -
127
0.74-0.38; P=0.40).
128
We next investigated the kinetics of AbNEI levels relative to that of CMV DNA load
129
during episodes of CMV DNAemia. A total of 110 sera (median, 7.5; Range, 2-24)
130
from 20 patients with CMV DNAemia were available for analysis. No correlation was
131
found between AbNEI levels and plasma CMV DNA loads (rho=0.09; 95% C.I. -0.52-
132
0.64; P=0.78). In fact, AbNEI levels remained constant in 14/22 episodes in the face of
133
decreasing CMV DNAemia levels. In the remaining 8 episodes, either decreasing (n=4),
134
fluctuating (n=3) or increasing levels of AbNEI levels were seen. The dynamics of 6
135
AbNEI levels in patients without CMV DNAemia was similar (stable levels in 11/13
136
patients). Likewise, CMV IgG levels were not correlated with CMV DNA loads within
137
episodes of active CMV infection (rho=0.19; 95% C.I. -0.28-0.74; P=0.68).
138
CMV-specific polyfunctional CD8+ T cells modulate the rate of CMV replication
139
within episodes of active CMV infection. T-cell immunity data were available from 29
140
out of the 39 patients included in the cohort, of whom 19 had CMV DNAemia within
141
the study period. Seven out of these 19 patients developed CMV DNAemia prior to the
142
first immunological monitoring timepoint and were excluded from the analyses reported
143
below. The data for each patient are shown in Table 1, and representative flow
144
cytometry plots of these patients are depicted in Figure 4. The prevalence of detectable
145
polyfunctional CD8+ T-cell responses (any combination of bifunctional T cells and/or
146
trifunctional T cells) was found to be comparable in patients with or without subsequent
147
CMV DNAemia (8/12 in patients with CMV DNAemia and 6/10 in patients without
148
CMV DNAemia; P=0.87). CMV-specific polyfunctional CD8+ T- cell levels did not
149
differ significantly between patients with or without subsequent CMV DNAemia
150
(median, 0.31 cell/μl; Range, 0.02-2.60 cells/μl, and median, 0.62 cell/μl; Range, 0.02-
151
2.01 cells/μl, respectively; P=0.91); Nevertheless, as shown in Table 3, patients
152
displaying detectable polyfunctional CD8+ T-cell responses at day +30 after Allo-SCT
153
had lower initial and peak CMV DNA loads in subsequent episodes of CMV DNAemia
154
than those who did not, although a statistical significance was not reached. Furthermore,
155
episodes of CMV DNAemia in the latter patients lasted significantly longer than those
156
in the former patients.
157
DISCUSSION 7
158
To our knowledge this is the first study assessing the potential role of CMV-specific
159
AbNEI in the control of CMV infection in the Allo-SCT setting. Prior studies
160
measuring CMV-specific antibodies neutralizing fibroblast infection failed to provide
161
convincing evidence of a role for these antibodies in either the prevention or clearance
162
of CMV viremia (Muñoz et al., 2001; Schoppel et al., 1998; Volpi et al., 1999); Here,
163
we performed a neutralization assay in which neutralizing antibodies likely target
164
epitopes within one or more of the components of the glycoprotein complex gH–gL–
165
pUL131A–pUL130–pUL128 or even epitopes on gB that are specifically involved in
166
pH-dependent triggering and/or with interactions between gB and the pentameric
167
glycoprotein complex. Our data suggested that CMV-specific AbNEI are unlikely to
168
play a major role in either the prevention or the resolution of CMV DNAemia. Firstly,
169
the dynamic pattern of AbNEI levels in patients who later developed CMV DNAemia
170
was indistinguishable from that seen in patients who did not. Secondly, the kinetics and
171
the level of CMV replication within episodes of CMV DNAemia was independent of
172
AbNEI levels measured prior to the occurrence of CMV DNAemia. Thirdly, no
173
apparent relationship between CMV DNA load kinetics and AbNEI level dynamics was
174
observed within episodes of CMV DNAemia. In fact, stable AbNEI levels were
175
frequently seen in concomitance with decreasing CMV DNA load levels.
176
Strikingly, patients exhibiting high baseline and peak AbNEI levels were more likely to
177
develop CMV DNAemia. In fact, a cut-off AbNEI level was established (4.7 log2 titer)
178
that discriminated fairly well (specificity and sensitivity ≥80%) between patients with or
179
without subsequent CMV DNAemia. In contrast, the levels of AD169-specific IgGs
180
measured by CLIA were not predictive of the occurrence of CMV DNAemia. In our
181
opinion, should AbNEI display a relevant role in conferring protection against CMV 8
182
viremia, the opposite phenomenon would be expected to occur, as has been shown for
183
CMV-specific functional T cells (Solano et al., 2008; Tormo et al., 2010a,b; Tormo et
184
al., 2011). Several non-mutually exclusive explanations may account for this finding.
185
First, CMV reactivation at tissue or mucosal sites occurring either during conditioning,
186
or very early after transplant may have boosted residual memory B cells in the recipient,
187
or memory B cells transferred within the allograft (Yamazaki et al., 2014) to produce
188
antibodies with such functional properties. Second, baseline serum levels of AbNEI
189
may reflect latent viral load, which ultimately determines the overall risk of CMV
190
reactivation (Forster et al., 2010; Reddehase et al., 1994). Third, the possibility exists
191
that a fraction of the antibodies quantitated in the neutralization assay employed in the
192
current study may enhance CMV infectivity in this clinical setting. In this sense, CMV-
193
specific antibodies of low neutralizing activity have been shown to enhance human
194
placenta infection by allowing receptor-mediated transcytosis of CMV-IgG complexes
195
that retain infectivity by Fc receptors (Maidji et al., 2006).
196
New compounds exhibiting higher intrinsic activity against CMV and better safety
197
profile than (val)ganciclovir have been developed (Boeckh et al., 2015). This has
198
renewed interest in antiviral prophylaxis for the prevention of CMV-related morbidity
199
in the Allo-SCT recipient (Chemaly et al., 2014). The identification of biomarkers
200
predicting the occurrence of CMV viremia (particularly those episodes requiring the
201
administration of preemptive antiviral therapy) would allow the inception of antiviral
202
prophylaxis on an individual basis according to the patient’s risk (targeted prophylaxis).
203
In this context, quantitation of AbNEIs may serve to that purpose.
9
204
There is increasing evidence to show that polyfunctional T cells are superior to those
205
exhibiting monofunctional properties in controlling acute and chronic viral infections,
206
such as those caused by HIV and HCV (Harari et al., 2006). Nevertheless, little is
207
known about the relevance of polyfunctional CD8+ T in the control of CMV infection in
208
the Allo-SCT setting. Preliminary evidence suggest that these cells may exert a major
209
role in the control of CMV replication episodes (Lacey et al., 2006; Lilleri et al., 2008;
210
Zhou et al., 2008; Król et al, 2011; Muñoz-Cobo et al., 2012). Data on CMV (pp65/IE-
211
1)-specific CD8+ T-cell immunity were available from a number of patients who either
212
did or did not develop a subsequent episode of CMV DNAemia. In our clinical setting,
213
the occurrence of CMV DNAemia could not be reliably predicted on the basis of either
214
the presence or absence of detectable polyfunctional CD8+ T-cell responses or the total
215
number of CMV-specific polyfunctional CD8+ T cells; Nevertheless, the mere presence
216
of these functional T cell types was associated with lower levels of CMV replication
217
within episodes of active CMV infection, and hence with a shorter duration of episodes.
218
This is in line with previously published data (Lilleri et al., 2008; Zhou et al., 2009;
219
Muñoz-Cobo et al., 2012), and further highlights the crucial role of adaptive T-cell
220
immunity in the control of CMV viremia in Allo-SCT recipients.
221
The apparent lack of contribution of CMV-specific AbNEI to the control of CMV
222
infection in CMV-seropositive Allo-SCT setting reported herein is in contrast to that
223
seen in the context of primary CMV infection during pregnancy (Lilleri et al., 2012;
224
Lilleri et al., 2013). In this latter setting, fetal CMV transmission was found to correlate
225
with delayed generation of maternal antibodies to gH/gLpUL128-130-131 complex
226
during primary infection. Taken together these data point to a more efficient role of
227
AbNEI in the control of CMV replication in the immunocompetent CMV-naïve host 10
228
than in that occurring in chronically infected immunosuppressed individuals as a result
229
of virus reactivation or reinfection. Further studies are needed to shed light on this issue.
230
It is being increasingly appreciated that immune mechanisms exerting control over
231
CMV infection are possibly redundant (Solano&Navarro, 2010). In this sense, data
232
obtained in the murine mCMV model indicated that adoptive transfer of memory B cells
233
from immune animals protected T-cell- and B-cell-deficient RAG-1(-/-) mice from
234
lethal challenge. This strategy was also effective in protecting from an already ongoing
235
viral infection (Klenovsek et al., 2007). In this context, it is of note that almost 50% of
236
patients in our cohort exhibited detectable CMV-specific polyfunctional CD8+ T-cell
237
responses. Hence, our conclusions must be taken in the context of our particular group
238
of patients and should not be extrapolated to patients displaying distinct immunological
239
backgrounds.
240
In summary, our data argue against a major role of CMV-specific AbNEI in the control
241
of CMV infection in the Allo-SCT setting, while reinforce the relevance of T cell
242
immunity for that purpose. Further studies involving larger cohorts are nevertheless
243
needed to either confirm or refute our observations.
244
PATIENTS AND METHODS
245
Patients. The current cohort consisted of 39 non-consecutive CMV-seropositive Allo-
246
SCT recipients (20 males and 19 females) undergoing Allo-SCT for hematological
247
malignancy at the Hospital Clínico Universitario of Valencia between 2010 and 2013.
248
The median age of patients was 52 years (range, 25 to 66 years). Allo-SCT (peripheral
249
blood, n=34; umbilical cord blood, n=4; bone marrow, n=1) for hematological
11
250
malignancy at the Hospital Clínico Universitario of Valencia between 2010 and 2013.
251
The median age of patients was 52 years (range, 25 to 66 years). Donors were HLA-
252
matched (71.8%) or HLA-mismatched (28.9%). The conditioning regimen was non-
253
myeloablative in 71.8% of transplants. Donors were CMV-seropositive in 23 cases and
254
CMV-seronegative in the remaining 16 cases. No patients received i.v immunoglobulins
255
during the study. The study period comprised the first 100 days following
256
transplantation. Only initial episodes of CMV DNAemia were considered for the
257
analyses reported herein. The study was approved by the local Review Board and Ethics
258
Committee. All patients gave written informed consent prior to participation in the
259
study.
260
Management of CMV infection. Monitoring of plasma CMV DNAemia was
261
performed once a week by using a real-time PCR assay (CMV real-time PCR or Abbott
262
RealTime CMV, Abbott Molecular, Des Plaines, IL, USA) (Clari et al., 2013; Gimeno
263
et al., 2008; Giménez et al., 2014). Patients were treated preemptively with antivirals as
264
described previously (Solano et al., 2008; Tormo et al., 2010a,b; Tormo et al., 2011).
265
CMV DNAemia (active CMV infection) was defined by the detection of any level of
266
CMV DNA in plasma. The duration of a given episode of CMV DNAemia was that
267
comprised between the day of the first positive PCR result and the day of the first
268
negative (undetectable) result.
269
Cells and virus. Human ARPE-19 retinal pigment epithelial cells (ATCC CRL-2302)
270
were cultured in high glucose Dulbecco’s modified Eagle medium-DMEM- (Gibco-
271
BRL) supplemented with 10% fetal calf serum (HyClone Laboratories), 10,000 IU/L
272
penicillin, 10 mg/L streptomycin (Gibco-BRL). CMV strain BADrUL131-Y4 was 12
273
kindly provided by Dr. T. Shenk (Princeton University, USA). This strain is derived
274
from a bacterial artificial chromosome clone of the CMV strain AD169 genome that had
275
been modified in Escherichia coli to express a functional UL131 protein, which permits
276
replication in either ARPE-19 or MRC-5 cells (Hahn et al., 2003; Wang & Shenk,
277
2005).Viral titers of (BADrUL131-Y4) were determined by limiting dilution in 96-well
278
plates using MRC-5 cells.
279
Neutralization assay. The neutralization assay was performed following a previously
280
published protocol with some modifications (Cui et al., 2008). Heat-inactivated sera
281
were serially diluted (twofold) in DMEM (from 1/5 to 1/2560). Fifty μl of the respective
282
serum dilution was incubated in triplicate (for 2 h at 37 °C) with the BADrUL131-Y4
283
virus inoculum (0.5 M.O.I in 50 μl of DMEM). Virus–antibody mixtures were then
284
added to confluent ARPE-19 monolayers previously seeded in 96-well microtiter plates
285
(clear/flat-bottomed 96-well plates-Costar). Infected and mock infected cells were used
286
as positive and negative controls, respectively. After 48 h of incubation at 37 °C in a 5%
287
CO2 atmosphere, cells were washed and fixed with cold 1:4 PBS:acetone solution. Cells
288
were stained with a p72-(immediate-early-1 protein) specific mAb (MAB810R CMV,
289
Clone 8B1.2, Millipore) diluted 1:2000 and washed with PBS+0.05% Tween-20. An
290
HRP-conjugated IgG anti-mouse secondary antibody (Sigma-Aldrich) diluted 1:1000
291
was next added. One-hundred l of peroxidase substrate (3,3´,5,5´-Tetramethyl-
292
benzidine Liquid Substrate, Supersensitive, for ELISA; Sigma-Aldrich) were added to
293
the mixture and incubated for 15 minutes. The reaction was stopped with 50 l of 1N
294
sulfuric acid and the absorbance was measured in a Microplate Reader with a 450 - 490
295
nm filter. The mean (triplicates) serum dilution inhibiting virus infectivity by 50% or 13
296
more with respect to infected cell controls was reported as the neutralizing-antibody
297
titer. Sera were scheduled to be obtained on a weekly basis. Variation in antibody titers
298
>1 log2 were deemed to be significant. A total of 338 sera were subjected to analysis.
299
CMV antibody assay. CMV (AD169)-specific IgG antibodies were measured by the
300
Chemiluminescent (CLIA) Architect CMV IgG assay (Abbott, IL, USA) following the
301
recommendations of the manufacturer. IgG antibody levels >6 absorbance units
302
(AU)/ml were considered reactive. Variations in CMV IgG levels >20% (the interassay
303
coefficient of variation-as reported by the manufacturer-+2SD) were considered to be
304
significant.
305
Enumeration of polyfunctional CMV-specific CD8+ T cells. Quantitation of pp65 and
306
IE-1-specific monofunctional (IFN-γ, TNF-α, and CD107a), bifunctional (IFN-γ/ TNF-
307
α, IFN-γ/CD107a, and TNF-α/CD107a) and trifunctional (IFN-γ/TNF-α/CD107a)
308
CMV-specific CD8+ T cells was performed by flow cytometry for intracellular cytokine
309
staining (ICS) at day +30 after transplant, as previously described (Muñoz-Cobo et al.,
310
2012). Briefly, whole blood samples (0.5 mL) were simultaneously stimulated with two
311
sets of 15-mer overlapping peptides encompassing the sequence of pp65 and IE-1 CMV
312
proteins (1 μg/mL/peptide), obtained from JPT peptide Technologies GmbH (Berlin,
313
Germany), in the presence of 1μg/mL of co-stimulatory mAbs to CD28 and CD49d, an
314
anti-CD107a mAb coupled to APC, brefeldin (5 µg/mL), and monensin (1 µM) for 6 h
315
at 37 ºC. The specimens were maintained at room temperature until stimulation, which
316
was performed within 24 h. after sampling. The cells were washed in PBS-2% FCS,
317
lysed in BD FACS lysis solution, stained with surface markers (anti-CD8-PerCP-Cy5-5
318
and anti-CD3-APC-Cy7), permeabilized (BD FACS Permeabilizing solution 2), 14
319
washed, and finally stained for intracellular cytokines (anti-IFN-γ-FITC and anti-TNF-
320
α-PE). All antibodies and solutions were purchased from Becton Dickinson (San Jose,
321
CA, USA). The cells were stored at 4 ºC in PBS-1% formaldehyde, acquired within 4 h
322
in a BD FACScantoII flow cytometer (BD Biosciences Immunocytometry Systems, San
323
Jose, CA) and analyzed with the program BD FACS Diva Software (BD Biosciences).
324
Negative controls (absence of peptide stimulation) were processed in parallel for all
325
experiments. After initial gating on lymphocytes, cells were selected on the basis of
326
CD3+ and CD8+ staining, and then further gated based on the expression of IFN-γ, TNF-
327
α, and CD107a (see Figure 4). Boolean gating analysis was performed to enumerate the
328
frequencies of all possible combinations. The total number of each CD8+ T cell
329
subpopulation was calculated by multiplying the corresponding percentage of CMV-
330
specific T cells (after background subtraction) by the absolute number of CD8+ T cells.
331
Specific (detectable) responses were considered as those that were >0.1% (mean+2
332
standard deviation-SD- of CMV-seronegative controls) for IFN-γ, TNF-α, and CD107a-
333
expressing CD8+ T-cell populations. Polyfunctional CD8+ T cells were those that were
334
positive for 2 (bifunctional) or 3 markers (trifunctional).
335
Statistical analysis. The data were analyzed with the aid of the statistical package SPSS
336
version 20.0 (SPSS, North Chicago, IL). Differences between medians were compared
337
using the Mann-Whitney U-test. The Spearman’s rank test was used to analyze
338
correlations between continuous variables. Two-sided exact P values are reported. A P
339
value 1 available specimens are
516
shown.
517
Fig 2. Baseline and peak cytomegalovirus (CMV)-specific antibodies
518
neutralizing epithelial infection (AbNEI) titers (panels A and C) and CMV IgGs
519
as measured by CLIA (panels C and D) in allogeneic stem cell transplant
520
recipients (Allo-SCT) who did or did not develop a subsequent episode of CMV
521
DNAemia. Bars represent median titers and the range of antibody titers. Two-
522
sided exact P values are shown. 23
523
Fig. 3. Receiver operating characteristics (ROC) curve analysis for
524
cytomegalovirus-specific antibodies neutralizing epithelial infection (ROC
525
curves for baseline and peak antibody titers were similar) indicating the optimal
526
cutoff value (4.7 log2 titer) for discriminating between patients with or without
527
subsequent CMV DNAemia (sensitivity of 83% and a specificity of 80%). Dots
528
represent different log2 antibody titers.
529
Fig. 4. Gating strategy for enumeration of CMV pp65 and IE-1-specific
530
polyfunctional CD8+ T-cell responses and representative flow cytometry plots
531
(FCP). FCP belonging to four different patients included in the cohort (patients
532
20, 11 and 33 in Table 1) are shown. Blood specimens of these patients were
533
stimulated, stained and acquired in different days. Cells were gated on
534
forward/side scatter characteristics do identify viable lymphocytes (not shown),
535
followed by gating on CD3+/CD8+ cells (row A), CD8+/CD107a (row B),
536
CD8+ /IFN-γ (row D) and CD8+ /TNFα (not shown). CD8+/CD107a
537
expressing T cells were further analyzed for their expression of TNF-α and
538
IFN-γ (row C). In turn, CD8+ /IFN-γ T cells were analyzed for their expression
539
of TNF-α and CD107a (row E).
540 541 542 543 544 545 24
Table 1. Relevant clinical, serological and immunological data from allogeneic stem cell transplant recipients included in the study CMV peak load CMV serostatus Patient no.
CMV DNAemia
Antiviral
(copies/mL)/ CMV
therapy
DNAemia (duration in
Log2 Baseline AbNEI
Polyfunctional
titer/Peak AbNEI titer *
CD8+ T cells (cells/μL) †
days)
25
1
D+R+
Yes
Yes
1062 / 45
7.85 / 7.85
NA
2
D-R+
Yes
Yes
13165 / 40
7.85/ 7.85
NA
3
D+R+
Yes
Yes
3656 / 89
5.77 / 5.77
0.00 ††
4
D+R+
Yes
Yes
40851 / 73
5.77 / 5.77
5.83 ††
5
D-/R+
Yes
Yes
1108 / 57
3.69 /4.38
0.00 ††
6
D+R+
Yes
Yes
17270 / 150
NA/ NA
0.00 ††
7
D+R+
Yes
Yes
2842 / 36
7.15 / 7.15
1.10
26
8
D-/R+
Yes
Yes
1455 / 142
5.77 / 5.77
0.06
9
D-/R+
Yes
Yes
2039 / 124
5.77 / 5.77
0.23
10
D-/R+
Yes
Yes
4873 / 127
6.46 / 6.46
0.15
11
D+/R-
Yes
Yes
4828 / 30
2.30 / 3.00
0.00
12
D-/R+
Yes
Yes
2094 / 34
5.77 / 6.46
0.00
13
D-R+
Yes
No
272 / 115
6.46 / 6.46
NA
14
D+R+
Yes
No
725 / 64
7.15 / 7.15
NA
15
D+R+
Yes
No
94 / 19
NA / NA
NA
16
D+R+
Yes
No
106 / 23
5.08 / 5.08
2.49 ††
17
D+R+
Yes
No
74 / 44
6.46 / 7.15
1.66 ††
18
D-/R+
Yes
No
277 / 49
3.69 / 3.69
0.00 ††
19
D+R+
Yes
No
144 / 67
7.15 / 7.15
0.00
20
D+R+
Yes
No
53 / 18
7.15 / 7.85
0.94
27
21
D+R+
Yes
No
593 / 36
6.46 / 6.46
0.31
22
D-/R+
Yes
No
60 / 7
7.15 / 7.15
0.02
23
D+R+
Yes
No
53 / 7
4.38 / 4.38
1.49
24
D+R+
Yes
No
907 / 41
5.08 / 5.08
0.00
25
D+R+
No
NA
NA
6.46 / 6.46
NA
26
D+R+
No
NA
NA
5.08 / 5.08
NA
27
D+R+
No
NA
NA
3.00 / 3.00
NA
28
D-/R+
No
NA
NA
1.61 / 5.08
NA
29
D-/R+
No
NA
NA
1.61 / 1.61
NA
30
D+R+
No
NA
NA
3.69 / 4.38
0.79
31
D-/R+
No
NA
NA
7.85 / 7.85
0.74
32
D-/R+
No
NA
NA
4.38 / 4.38
0.09
33
D+R+
No
NA
NA
2.30 / 3.69
0.02
34
D-/R+
No
NA
NA
1.61 / 3.00
0.49
35
D+R+
No
NA
NA
4.38 / 4.38
2.01
36
D+R+
No
NA
NA
3.69 / 4.38
0.00
37
D-/R+
No
NA
NA
1.61 / 3.00
0.00
38
D+R+
No
NA
NA
1.61 / 1.61
0.00
39
D-/R+
No
NA
NA
1.61 / 1.61
0.00
Abbreviations: AbNEI, antibodies neutralizing epitelial infection; CMV, Cytomegalovirus; D, Donor; NA, not applicable; R, Recipient. *
AbNEI titer (inverse) in patients with or without subsequent CMV DNAemia
† Total cell count of polyfunctionals CD8+ T cells (expressing 2 or more of the following markers: IFN-γ/TNF-α, IFN- γ/CD107a) at day +30 after transplant as determined by flow cytometry for intracellular cytokine staining. †† Patients with CMV DNAemia prior the time of immunological monitoring. 546 547 28
548 549 Table 2. Kinetics of CMV DNAemia in patients displaying baseline and peak cytomegalovirus-specific antibodies neutralizing epithelial infection (AbNEI) either below or above 4.7 log2 titer. Baseline AbNEI Parameter
Median initial CMV DNA load in copies/ml (range) Median peak CMV DNA load in copies/ml (range) Median duration of CMV DNAemia in days (range)
log2 titer (inverse)
Peak AbNEI log2 titer P
(inverse)
value* < 4.7
> 4.7
52 (16-
54 (10-
91)
188)
693 (53-
985 (53-
4828)
40851)
39.5 (7-
44.5 (7-
57)
150)
0.76
0.67
0.30
P value*
< 4.7
> 4.7
51 (16-
55 (10-
91)
188)
1108 (53-
907 (53-
17270)
40851)
49 (7-
44 (7-
150)
142)
0.41
0.72
0.89
*Differences between medians were compared using the Mann-Whitney U-test. Twosided exact P values are reported. A P value 0.1% (mean+2 standard deviation-SD- of CMVseronegative controls) after background substraction for IFN-γ, TNF-α, and CD107aexpressing CD8+ T-cell populations. †Differences between medians were compared using the Mann-Whitney U-test. Twosided exact P values are reported. A P value
Journal of General Virology The role of cytomegalovirus (CMV)-specific polyfunctional CD8+ T cells and antibodies neutralizing virus epithelial infection in the control of CMV infection in the allogeneic stem cell transplantation setting. --Manuscript Draft-Manuscript Number:
VIR-D-15-00162R2
Full Title:
The role of cytomegalovirus (CMV)-specific polyfunctional CD8+ T cells and antibodies neutralizing virus epithelial infection in the control of CMV infection in the allogeneic stem cell transplantation setting.
Short Title:
CMV-specific neutralizing antibodies and CMV DNAemia
Article Type:
Standard
Section/Category:
Animal - Large DNA Viruses
Corresponding Author:
David Navarro School of Medicine, University of Valencia, Spain Valencia, SPAIN
First Author:
Estela Giménez
Order of Authors:
Estela Giménez Pilar Blanco-Lobo Beatriz Muñoz-Cobo Carlos Solano Paula Amat Pilar Pérez-Romero David Navarro
Abstract:
The role cytomegalovirus (CMV)-specific polyfunctional CD8+ T cells and that of antibodies neutralizing virus epithelial infection (AbNEI) in the control of CMV DNAemia was investigated in 39 CMV-seropositive allogeneic stem cell transplant recipients (Allo-SCT) with (n=24) or without (n=15) CMV DNAemia. AbNEI levels were prospectively monitored by means of a neutralization assay employing retinal epithelial cells (ARPE-19) and the recombinant CMV strain BADrUL131-Y4. Quantification of CMV-specific polyfunctional CD8+ T cells (expressing 2 or 3 of the following markers: IFN-γ, TNF-α, CD107a) in whole blood was performed by flow cytometry for intracellular cytokine staining. We found no differences in the dynamic pattern of AbNEI in patients with or without subsequent CMV DNAemia. Baseline and peak AbNEI titers were not predictive of the dynamics of CMV replication within episodes. No correlation was found between CMV DNA loads and AbNEI levels during episodes of CMV DNAemia (rho=0.09; 95% C.I. -0.52-0.64; P=0.78). The detection of pp65/IE-1 CMVspecific polyfunctional CD8+ T cells was associated with low level viral replication within subsequent episodes of CMV DNAemia. Interestingly, the presence of AbNEI titers (inverse) >4.7 log2 was predictive of the occurrence of CMV DNAemia (sensitivity, 83%; Specificity, 80%). Our findings should add insights to the role of humoral and cellular immunity in the control of CMV infection in the Allo-SCT setting.
Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Manuscript Including References (Word document) Click here to download Manuscript Including References (Word document): PP DN JGV 2015 text tables revised.2 definitive.doc
1
VIR-D-15-00162.R2
2
The role of cytomegalovirus (CMV)-specific polyfunctional CD8+ T cells and
3
antibodies neutralizing virus epithelial infection in the control of CMV infection in
4
the allogeneic stem cell transplantation setting
5
Estela Giménez,1# Pilar Blanco-Lobo,2# Beatriz Muñoz-Cobo,1 Carlos Solano,3 Paula
6
Amat,3 Pilar Pérez-Romero,2 and David Navarro1,4
7
1
8
Spain.
9
2
Microbiology Service, Hospital Clínico Universitario, Fundación INCLIVA, Valencia,
Instituto
de
Biomedicina
de
Sevilla,
Hospital
Universitario
Virgen
del
10
Rocío/CSIC/Universidad de Sevilla, Sevilla, Spain.
11
3
12
INCLIVA, Valencia, Spain.
13
4
14
Spain.
15
#Authors contributed equally to this work.
16
Author for correspondence: David Navarro. Microbiology Service, Hospital Clínico
17
Universitario, and Department of Microbiology, School of Medicine, Valencia, Spain.
18
Av. Blasco Ibáñez 17, 46010 Valencia, Spain. Phone: 34(96)3864657; Fax:
19
34(96)3864173; E-mail: [email protected]
20
Running Title: CMV-specific neutralizing antibodies and CMV DNAemia.
Hematology and Medical Oncology Service, Hospital Clínico Universitario, Fundación
Department of Microbiology, School of Medicine, University of Valencia, Valencia,
1
21
Word counts. Summary: 216; Text: 3,541.
22
SUMMARY
23
The role cytomegalovirus (CMV)-specific polyfunctional CD8+ T cells and that of
24
antibodies neutralizing virus epithelial infection (AbNEI) in the control of CMV
25
DNAemia was investigated in 39 CMV-seropositive allogeneic stem cell transplant
26
recipients (Allo-SCT) with (n=24) or without (n=15) CMV DNAemia. AbNEI levels
27
were prospectively monitored by means of a neutralization assay employing retinal
28
epithelial cells (ARPE-19) and the recombinant CMV strain BADrUL131-Y4.
29
Quantification of CMV-specific polyfunctional CD8+ T cells (expressing 2 or 3 of the
30
following markers: IFN-γ, TNF-α, CD107a) in whole blood was performed by flow
31
cytometry for intracellular cytokine staining. We found no differences in the dynamic
32
pattern of AbNEI in patients with or without subsequent CMV DNAemia. Baseline and
33
peak AbNEI titers were not predictive of the dynamics of CMV replication within
34
episodes. No correlation was found between CMV DNA loads and AbNEI levels during
35
episodes of CMV DNAemia (rho=0.09; 95% C.I. -0.52-0.64; P=0.78). The detection of
36
pp65/IE-1 CMV-specific polyfunctional CD8+ T cells was associated with low level
37
viral replication within subsequent episodes of CMV DNAemia. Interestingly, the
38
presence of AbNEI titers (inverse) >4.7 log2 was predictive of the occurrence of CMV
39
DNAemia (sensitivity, 83%; Specificity, 80%). Our findings should add insights to the
40
role of humoral and cellular immunity in the control of CMV infection in the Allo-SCT
41
setting.
2
42
Key Words: Cytomegalovirus (CMV), CMV DNAemia, Allogeneic stem cell
43
transplantation (Allo-SCT), neutralizing antibodies, CMV-specific polyfunctional CD8+
44
T cells.
45
INTRODUCTION
46
Protection from and clearance of cytomegalovirus (CMV) viremia in allogeneic stem
47
cell transplant recipients (Allo-SCT) is critically dependent upon the reconstitution and
48
adequate expansion, respectively, of functional CMV-specific T cells (Solano &
49
Navarro, 2010). Whether CMV-specific antibodies exerting antiviral activities
50
contribute in a relevant way to these purposes has not been elucidated. In this context,
51
the
52
hyperimmunoglobulin on minimizing the incidence and fatal outcome of CMV end-
53
organ disease, claimed in this clinical setting, has never been conclusively proven
54
(Guglielmo et al., 1994). Furthermore, contradictory data have been published on the
55
role of CMV-specific neutralizing antibodies in the prevention or control of CMV
56
viremia (Muñoz et al., 2001; Schoppel et al., 1998; Volpi et al., 1999); In these studies,
57
neutralizing antibodies were quantified by means of a classical neutralization assay
58
involving the use of the CMV AD169 strain and human-derived fibroblasts, which
59
mostly measures antibody levels against gB and the glycoprotein complex gH–gL, both
60
essential for infection of human fibroblasts (Vanarsdall et al., 2008). Recently, the
61
virion glycoprotein complex gH–gL–pUL131A–pUL130–pUL128 has been shown to
62
be required for infection of endothelial, epithelial and myeloid cells (Gerna et al., 2005;
63
Hahn et al., 2004; Ryckman et al., 2008a,b; Wang & Shenk, 2005a), all of which are
64
key targets for CMV pathogenesis in the human host. Neutralizing antibodies targeting 3
beneficial
effect
of
the
administration
of
CMV-specific
human
65
this pentameric complex are known to display a much higher potency in neutralizing
66
virus infection of epithelial and endothelial cells than those against gB (Macagno et al.,
67
2010). The AD169 strain of CMV carries a defective UL131 ORF, thus antibodies
68
against this pentameric complex are not measured in the classical neutralization assay
69
(Hahn et al., 2004). Antibodies against neutralizing epitopes of the complex gH–gL–
70
pUL131A–pUL130–pUL128 have been shown to correlate with virus control in
71
immunocompetent individuals (Lilleri et al., 2012) and protection from fetal CMV
72
transmission (Lilleri et al., 2013). Here, we investigated in parallel the role of CMV-
73
specific antibodies that neutralize epithelial cell infection (AbNEI) and that of
74
cytomegalovirus (CMV)-specific polyfunctional CD8+ T in the control of CMV
75
DNAemia in a cohort of Allo-SCT recipients.
76
RESULTS
77
Incidence of CMV DNAemia. A total of 24 out of 39 patients (61.5%) developed a
78
first episode of CMV DNAemia within the first 100 days after Allo-SCT (median 30
79
days; Range 26-42 days). All episodes were eventually controlled. Twelve episodes
80
required the administration of pre-emptive antiviral therapy. CMV DNAemia was
81
spontaneously cleared in the remaining 12 patients.
82
Kinetics of CMV-specific antibodies neutralizing epithelial infection (AbNEI) in
83
patients with or without CMV DNAemia. We first assessed the kinetics of AbNEI in
84
patients with or without subsequent CMV DNAemia. Sequential sera (n=60) were
85
available from 18 patients subsequently having CMV DNAemia (median, 3/patient;
86
Range, 2-8/patient). Serial sera (n=76) from 15 patients with no CMV DNAemia
87
(median, 5/patient; Range, 3-8/patient) obtained within a comparable time frame were 4
88
also available for analysis and served as controls. The kinetic profiles of AbNEI did not
89
differ notably between patients in the study groups (Figure 1, panels A and B). CMV-
90
specific AbNEI levels were found to remain stable in 13/18 and 10/15 patients with or
91
without subsequent CMV DNAemia, respectively (P=0.73). Either increasing,
92
decreasing or fluctuating levels of AbNEI levels were observed at comparable
93
frequencies among individuals belonging to both groups (P=>0.5). Data on CMV CMV
94
(AD169)-specific IgG levels, as measured by CLIA, were available from 31 patients (a
95
median of 3 samples/patient; Range, 1-6). All sera having detectable AbNEI were also
96
reactive by the CMV IgG CLIA. As shown in Figure 1 (panels C and D), the kinetics
97
pattern of CMV IgG antibodies was found to be comparable in patients with or without
98
subsequent CMV DNAemia, with a rather similar number of patients displaying stable,
99
increasing or decreasing antibody levels in both study groups.
100
Quantitation of CMV-specific antibodies neutralizing epithelial infection allows for
101
anticipation of the occurrence of CMV DNAemia. We next investigated whether
102
AbNEI levels measured either at the time of or early after Allo-SCT were indicative of
103
the risk of CMV DNAemia. The data for each patient of the cohort are shown in Table
104
1.We found that both baseline (median, day 0; Range, day -1 to 16) and peak (median,
105
day 14; Range, days 0-32) AbNEI levels were significantly higher in patients who went
106
on to develop CMV DNAemia than in those who did not (Figure 2, panels A and B,
107
respectively). ROC curve analyses indicated that a cut-off of 4.7 log2 titer (1/120) for
108
both baseline and peak AbNEI discriminated fairly well (sensitivity, 83%; specificity,
109
80%) between patients with or without subsequent CMV DNAemia (Figure 3). In fact,
110
the odds ratio for developing CMV DNAemia was 20.0 (95% C.I 3.81-105.11) for
111
patients with AbNEI titers above the referred threshold. In contrast, both baseline and 5
112
peak CMV-specific IgG titers were comparable in both study groups (Figure 2, panels C
113
and D). Thus, a CMV IgG threshold level discriminating between patients with or
114
without subsequent CMV DNAemia could not be established. Of note the fact that
115
patients displaying baseline and peak AbNEI titers either above or below this cut-off
116
were comparable in terms of pre-transplant and post-transplant clinical factors known to
117
modulate the risk of CMV DNAemia (not shown).
118
The dynamics of CMV DNAemia is not influenced by CMV-specific antibodies
119
neutralizing epithelial infection. We then investigated whether AbNEI levels
120
measured prior to the occurrence of CMV DNAemia allowed for the inference of the
121
dynamics of virus replication within episodes. As shown in Table 2, neither initial nor
122
peak plasma CMV DNA loads differed significantly between patients displaying
123
baseline or peak AbNEI levels above or below 4.7 log2 titer. Likewise, both the need for
124
pre-emptive antiviral therapy and the duration of episodes of CMV DNAemia were not
125
associated with AbNEI levels. Furthermore, no significant correlation was observed
126
between peak AbNEI levels and peak plasma CMV DNA loads (rho=-0.27; 95% C.I., -
127
0.74-0.38; P=0.40).
128
We next investigated the kinetics of AbNEI levels relative to that of CMV DNA load
129
during episodes of CMV DNAemia. A total of 110 sera (median, 7.5; Range, 2-24)
130
from 20 patients with CMV DNAemia were available for analysis. No correlation was
131
found between AbNEI levels and plasma CMV DNA loads (rho=0.09; 95% C.I. -0.52-
132
0.64; P=0.78). In fact, AbNEI levels remained constant in 14/22 episodes in the face of
133
decreasing CMV DNAemia levels. In the remaining 8 episodes, either decreasing (n=4),
134
fluctuating (n=3) or increasing levels of AbNEI levels were seen. The dynamics of 6
135
AbNEI levels in patients without CMV DNAemia was similar (stable levels in 11/13
136
patients). Likewise, CMV IgG levels were not correlated with CMV DNA loads within
137
episodes of active CMV infection (rho=0.19; 95% C.I. -0.28-0.74; P=0.68).
138
CMV-specific polyfunctional CD8+ T cells modulate the rate of CMV replication
139
within episodes of active CMV infection. T-cell immunity data were available from 29
140
out of the 39 patients included in the cohort, of whom 19 had CMV DNAemia within
141
the study period. Seven out of these 19 patients developed CMV DNAemia prior to the
142
first immunological monitoring timepoint and were excluded from the analyses reported
143
below. The data for each patient are shown in Table 1, and representative flow
144
cytometry plots of these patients are depicted in Figure 4. The prevalence of detectable
145
polyfunctional CD8+ T-cell responses (any combination of bifunctional T cells and/or
146
trifunctional T cells) was found to be comparable in patients with or without subsequent
147
CMV DNAemia (8/12 in patients with CMV DNAemia and 6/10 in patients without
148
CMV DNAemia; P=0.87). CMV-specific polyfunctional CD8+ T- cell levels did not
149
differ significantly between patients with or without subsequent CMV DNAemia
150
(median, 0.31 cell/μl; Range, 0.02-2.60 cells/μl, and median, 0.62 cell/μl; Range, 0.02-
151
2.01 cells/μl, respectively; P=0.91); Nevertheless, as shown in Table 3, patients
152
displaying detectable polyfunctional CD8+ T-cell responses at day +30 after Allo-SCT
153
had lower initial and peak CMV DNA loads in subsequent episodes of CMV DNAemia
154
than those who did not, although a statistical significance was not reached. Furthermore,
155
episodes of CMV DNAemia in the latter patients lasted significantly longer than those
156
in the former patients.
157
DISCUSSION 7
158
To our knowledge this is the first study assessing the potential role of CMV-specific
159
AbNEI in the control of CMV infection in the Allo-SCT setting. Prior studies
160
measuring CMV-specific antibodies neutralizing fibroblast infection failed to provide
161
convincing evidence of a role for these antibodies in either the prevention or clearance
162
of CMV viremia (Muñoz et al., 2001; Schoppel et al., 1998; Volpi et al., 1999); Here,
163
we performed a neutralization assay in which neutralizing antibodies likely target
164
epitopes within one or more of the components of the glycoprotein complex gH–gL–
165
pUL131A–pUL130–pUL128 or even epitopes on gB that are specifically involved in
166
pH-dependent triggering and/or with interactions between gB and the pentameric
167
glycoprotein complex. Our data suggested that CMV-specific AbNEI are unlikely to
168
play a major role in either the prevention or the resolution of CMV DNAemia. Firstly,
169
the dynamic pattern of AbNEI levels in patients who later developed CMV DNAemia
170
was indistinguishable from that seen in patients who did not. Secondly, the kinetics and
171
the level of CMV replication within episodes of CMV DNAemia was independent of
172
AbNEI levels measured prior to the occurrence of CMV DNAemia. Thirdly, no
173
apparent relationship between CMV DNA load kinetics and AbNEI level dynamics was
174
observed within episodes of CMV DNAemia. In fact, stable AbNEI levels were
175
frequently seen in concomitance with decreasing CMV DNA load levels.
176
Strikingly, patients exhibiting high baseline and peak AbNEI levels were more likely to
177
develop CMV DNAemia. In fact, a cut-off AbNEI level was established (4.7 log2 titer)
178
that discriminated fairly well (specificity and sensitivity ≥80%) between patients with or
179
without subsequent CMV DNAemia. In contrast, the levels of AD169-specific IgGs
180
measured by CLIA were not predictive of the occurrence of CMV DNAemia. In our
181
opinion, should AbNEI display a relevant role in conferring protection against CMV 8
182
viremia, the opposite phenomenon would be expected to occur, as has been shown for
183
CMV-specific functional T cells (Solano et al., 2008; Tormo et al., 2010a,b; Tormo et
184
al., 2011). Several non-mutually exclusive explanations may account for this finding.
185
First, CMV reactivation at tissue or mucosal sites occurring either during conditioning,
186
or very early after transplant may have boosted residual memory B cells in the recipient,
187
or memory B cells transferred within the allograft (Yamazaki et al., 2014) to produce
188
antibodies with such functional properties. Second, baseline serum levels of AbNEI
189
may reflect latent viral load, which ultimately determines the overall risk of CMV
190
reactivation (Forster et al., 2010; Reddehase et al., 1994). Third, the possibility exists
191
that a fraction of the antibodies quantitated in the neutralization assay employed in the
192
current study may enhance CMV infectivity in this clinical setting. In this sense, CMV-
193
specific antibodies of low neutralizing activity have been shown to enhance human
194
placenta infection by allowing receptor-mediated transcytosis of CMV-IgG complexes
195
that retain infectivity by Fc receptors (Maidji et al., 2006).
196
New compounds exhibiting higher intrinsic activity against CMV and better safety
197
profile than (val)ganciclovir have been developed (Boeckh et al., 2015). This has
198
renewed interest in antiviral prophylaxis for the prevention of CMV-related morbidity
199
in the Allo-SCT recipient (Chemaly et al., 2014). The identification of biomarkers
200
predicting the occurrence of CMV viremia (particularly those episodes requiring the
201
administration of preemptive antiviral therapy) would allow the inception of antiviral
202
prophylaxis on an individual basis according to the patient’s risk (targeted prophylaxis).
203
In this context, quantitation of AbNEIs may serve to that purpose.
9
204
There is increasing evidence to show that polyfunctional T cells are superior to those
205
exhibiting monofunctional properties in controlling acute and chronic viral infections,
206
such as those caused by HIV and HCV (Harari et al., 2006). Nevertheless, little is
207
known about the relevance of polyfunctional CD8+ T in the control of CMV infection in
208
the Allo-SCT setting. Preliminary evidence suggest that these cells may exert a major
209
role in the control of CMV replication episodes (Lacey et al., 2006; Lilleri et al., 2008;
210
Zhou et al., 2008; Król et al, 2011; Muñoz-Cobo et al., 2012). Data on CMV (pp65/IE-
211
1)-specific CD8+ T-cell immunity were available from a number of patients who either
212
did or did not develop a subsequent episode of CMV DNAemia. In our clinical setting,
213
the occurrence of CMV DNAemia could not be reliably predicted on the basis of either
214
the presence or absence of detectable polyfunctional CD8+ T-cell responses or the total
215
number of CMV-specific polyfunctional CD8+ T cells; Nevertheless, the mere presence
216
of these functional T cell types was associated with lower levels of CMV replication
217
within episodes of active CMV infection, and hence with a shorter duration of episodes.
218
This is in line with previously published data (Lilleri et al., 2008; Zhou et al., 2009;
219
Muñoz-Cobo et al., 2012), and further highlights the crucial role of adaptive T-cell
220
immunity in the control of CMV viremia in Allo-SCT recipients.
221
The apparent lack of contribution of CMV-specific AbNEI to the control of CMV
222
infection in CMV-seropositive Allo-SCT setting reported herein is in contrast to that
223
seen in the context of primary CMV infection during pregnancy (Lilleri et al., 2012;
224
Lilleri et al., 2013). In this latter setting, fetal CMV transmission was found to correlate
225
with delayed generation of maternal antibodies to gH/gLpUL128-130-131 complex
226
during primary infection. Taken together these data point to a more efficient role of
227
AbNEI in the control of CMV replication in the immunocompetent CMV-naïve host 10
228
than in that occurring in chronically infected immunosuppressed individuals as a result
229
of virus reactivation or reinfection. Further studies are needed to shed light on this issue.
230
It is being increasingly appreciated that immune mechanisms exerting control over
231
CMV infection are possibly redundant (Solano&Navarro, 2010). In this sense, data
232
obtained in the murine mCMV model indicated that adoptive transfer of memory B cells
233
from immune animals protected T-cell- and B-cell-deficient RAG-1(-/-) mice from
234
lethal challenge. This strategy was also effective in protecting from an already ongoing
235
viral infection (Klenovsek et al., 2007). In this context, it is of note that almost 50% of
236
patients in our cohort exhibited detectable CMV-specific polyfunctional CD8+ T-cell
237
responses. Hence, our conclusions must be taken in the context of our particular group
238
of patients and should not be extrapolated to patients displaying distinct immunological
239
backgrounds.
240
In summary, our data argue against a major role of CMV-specific AbNEI in the control
241
of CMV infection in the Allo-SCT setting, while reinforce the relevance of T cell
242
immunity for that purpose. Further studies involving larger cohorts are nevertheless
243
needed to either confirm or refute our observations.
244
PATIENTS AND METHODS
245
Patients. The current cohort consisted of 39 non-consecutive CMV-seropositive Allo-
246
SCT recipients (20 males and 19 females) undergoing Allo-SCT for hematological
247
malignancy at the Hospital Clínico Universitario of Valencia between 2010 and 2013.
248
The median age of patients was 52 years (range, 25 to 66 years). Allo-SCT (peripheral
249
blood, n=34; umbilical cord blood, n=4; bone marrow, n=1) for hematological
11
250
malignancy at the Hospital Clínico Universitario of Valencia between 2010 and 2013.
251
The median age of patients was 52 years (range, 25 to 66 years). Donors were HLA-
252
matched (71.8%) or HLA-mismatched (28.9%). The conditioning regimen was non-
253
myeloablative in 71.8% of transplants. Donors were CMV-seropositive in 23 cases and
254
CMV-seronegative in the remaining 16 cases. No patients received i.v immunoglobulins
255
during the study. The study period comprised the first 100 days following
256
transplantation. Only initial episodes of CMV DNAemia were considered for the
257
analyses reported herein. The study was approved by the local Review Board and Ethics
258
Committee. All patients gave written informed consent prior to participation in the
259
study.
260
Management of CMV infection. Monitoring of plasma CMV DNAemia was
261
performed once a week by using a real-time PCR assay (CMV real-time PCR or Abbott
262
RealTime CMV, Abbott Molecular, Des Plaines, IL, USA) (Clari et al., 2013; Gimeno
263
et al., 2008; Giménez et al., 2014). Patients were treated preemptively with antivirals as
264
described previously (Solano et al., 2008; Tormo et al., 2010a,b; Tormo et al., 2011).
265
CMV DNAemia (active CMV infection) was defined by the detection of any level of
266
CMV DNA in plasma. The duration of a given episode of CMV DNAemia was that
267
comprised between the day of the first positive PCR result and the day of the first
268
negative (undetectable) result.
269
Cells and virus. Human ARPE-19 retinal pigment epithelial cells (ATCC CRL-2302)
270
were cultured in high glucose Dulbecco’s modified Eagle medium-DMEM- (Gibco-
271
BRL) supplemented with 10% fetal calf serum (HyClone Laboratories), 10,000 IU/L
272
penicillin, 10 mg/L streptomycin (Gibco-BRL). CMV strain BADrUL131-Y4 was 12
273
kindly provided by Dr. T. Shenk (Princeton University, USA). This strain is derived
274
from a bacterial artificial chromosome clone of the CMV strain AD169 genome that had
275
been modified in Escherichia coli to express a functional UL131 protein, which permits
276
replication in either ARPE-19 or MRC-5 cells (Hahn et al., 2003; Wang & Shenk,
277
2005).Viral titers of (BADrUL131-Y4) were determined by limiting dilution in 96-well
278
plates using MRC-5 cells.
279
Neutralization assay. The neutralization assay was performed following a previously
280
published protocol with some modifications (Cui et al., 2008). Heat-inactivated sera
281
were serially diluted (twofold) in DMEM (from 1/5 to 1/2560). Fifty μl of the respective
282
serum dilution was incubated in triplicate (for 2 h at 37 °C) with the BADrUL131-Y4
283
virus inoculum (0.5 M.O.I in 50 μl of DMEM). Virus–antibody mixtures were then
284
added to confluent ARPE-19 monolayers previously seeded in 96-well microtiter plates
285
(clear/flat-bottomed 96-well plates-Costar). Infected and mock infected cells were used
286
as positive and negative controls, respectively. After 48 h of incubation at 37 °C in a 5%
287
CO2 atmosphere, cells were washed and fixed with cold 1:4 PBS:acetone solution. Cells
288
were stained with a p72-(immediate-early-1 protein) specific mAb (MAB810R CMV,
289
Clone 8B1.2, Millipore) diluted 1:2000 and washed with PBS+0.05% Tween-20. An
290
HRP-conjugated IgG anti-mouse secondary antibody (Sigma-Aldrich) diluted 1:1000
291
was next added. One-hundred l of peroxidase substrate (3,3´,5,5´-Tetramethyl-
292
benzidine Liquid Substrate, Supersensitive, for ELISA; Sigma-Aldrich) were added to
293
the mixture and incubated for 15 minutes. The reaction was stopped with 50 l of 1N
294
sulfuric acid and the absorbance was measured in a Microplate Reader with a 450 - 490
295
nm filter. The mean (triplicates) serum dilution inhibiting virus infectivity by 50% or 13
296
more with respect to infected cell controls was reported as the neutralizing-antibody
297
titer. Sera were scheduled to be obtained on a weekly basis. Variation in antibody titers
298
>1 log2 were deemed to be significant. A total of 338 sera were subjected to analysis.
299
CMV antibody assay. CMV (AD169)-specific IgG antibodies were measured by the
300
Chemiluminescent (CLIA) Architect CMV IgG assay (Abbott, IL, USA) following the
301
recommendations of the manufacturer. IgG antibody levels >6 absorbance units
302
(AU)/ml were considered reactive. Variations in CMV IgG levels >20% (the interassay
303
coefficient of variation-as reported by the manufacturer-+2SD) were considered to be
304
significant.
305
Enumeration of polyfunctional CMV-specific CD8+ T cells. Quantitation of pp65 and
306
IE-1-specific monofunctional (IFN-γ, TNF-α, and CD107a), bifunctional (IFN-γ/ TNF-
307
α, IFN-γ/CD107a, and TNF-α/CD107a) and trifunctional (IFN-γ/TNF-α/CD107a)
308
CMV-specific CD8+ T cells was performed by flow cytometry for intracellular cytokine
309
staining (ICS) at day +30 after transplant, as previously described (Muñoz-Cobo et al.,
310
2012). Briefly, whole blood samples (0.5 mL) were simultaneously stimulated with two
311
sets of 15-mer overlapping peptides encompassing the sequence of pp65 and IE-1 CMV
312
proteins (1 μg/mL/peptide), obtained from JPT peptide Technologies GmbH (Berlin,
313
Germany), in the presence of 1μg/mL of co-stimulatory mAbs to CD28 and CD49d, an
314
anti-CD107a mAb coupled to APC, brefeldin (5 µg/mL), and monensin (1 µM) for 6 h
315
at 37 ºC. The specimens were maintained at room temperature until stimulation, which
316
was performed within 24 h. after sampling. The cells were washed in PBS-2% FCS,
317
lysed in BD FACS lysis solution, stained with surface markers (anti-CD8-PerCP-Cy5-5
318
and anti-CD3-APC-Cy7), permeabilized (BD FACS Permeabilizing solution 2), 14
319
washed, and finally stained for intracellular cytokines (anti-IFN-γ-FITC and anti-TNF-
320
α-PE). All antibodies and solutions were purchased from Becton Dickinson (San Jose,
321
CA, USA). The cells were stored at 4 ºC in PBS-1% formaldehyde, acquired within 4 h
322
in a BD FACScantoII flow cytometer (BD Biosciences Immunocytometry Systems, San
323
Jose, CA) and analyzed with the program BD FACS Diva Software (BD Biosciences).
324
Negative controls (absence of peptide stimulation) were processed in parallel for all
325
experiments. After initial gating on lymphocytes, cells were selected on the basis of
326
CD3+ and CD8+ staining, and then further gated based on the expression of IFN-γ, TNF-
327
α, and CD107a (see Figure 4). Boolean gating analysis was performed to enumerate the
328
frequencies of all possible combinations. The total number of each CD8+ T cell
329
subpopulation was calculated by multiplying the corresponding percentage of CMV-
330
specific T cells (after background subtraction) by the absolute number of CD8+ T cells.
331
Specific (detectable) responses were considered as those that were >0.1% (mean+2
332
standard deviation-SD- of CMV-seronegative controls) for IFN-γ, TNF-α, and CD107a-
333
expressing CD8+ T-cell populations. Polyfunctional CD8+ T cells were those that were
334
positive for 2 (bifunctional) or 3 markers (trifunctional).
335
Statistical analysis. The data were analyzed with the aid of the statistical package SPSS
336
version 20.0 (SPSS, North Chicago, IL). Differences between medians were compared
337
using the Mann-Whitney U-test. The Spearman’s rank test was used to analyze
338
correlations between continuous variables. Two-sided exact P values are reported. A P
339
value 1 available specimens are
516
shown.
517
Fig 2. Baseline and peak cytomegalovirus (CMV)-specific antibodies
518
neutralizing epithelial infection (AbNEI) titers (panels A and C) and CMV IgGs
519
as measured by CLIA (panels C and D) in allogeneic stem cell transplant
520
recipients (Allo-SCT) who did or did not develop a subsequent episode of CMV
521
DNAemia. Bars represent median titers and the range of antibody titers. Two-
522
sided exact P values are shown. 23
523
Fig. 3. Receiver operating characteristics (ROC) curve analysis for
524
cytomegalovirus-specific antibodies neutralizing epithelial infection (ROC
525
curves for baseline and peak antibody titers were similar) indicating the optimal
526
cutoff value (4.7 log2 titer) for discriminating between patients with or without
527
subsequent CMV DNAemia (sensitivity of 83% and a specificity of 80%). Dots
528
represent different log2 antibody titers.
529
Fig. 4. Gating strategy for enumeration of CMV pp65 and IE-1-specific
530
polyfunctional CD8+ T-cell responses and representative flow cytometry plots
531
(FCP). FCP belonging to four different patients included in the cohort (patients
532
20, 11 and 33 in Table 1) are shown. Blood specimens of these patients were
533
stimulated, stained and acquired in different days. Cells were gated on
534
forward/side scatter characteristics do identify viable lymphocytes (not shown),
535
followed by gating on CD3+/CD8+ cells (row A), CD8+/CD107a (row B),
536
CD8+ /IFN-γ (row D) and CD8+ /TNFα (not shown). CD8+/CD107a
537
expressing T cells were further analyzed for their expression of TNF-α and
538
IFN-γ (row C). In turn, CD8+ /IFN-γ T cells were analyzed for their expression
539
of TNF-α and CD107a (row E).
540 541 542 543 544 545 24
Table 1. Relevant clinical, serological and immunological data from allogeneic stem cell transplant recipients included in the study CMV peak load CMV serostatus Patient no.
CMV DNAemia
Antiviral
(copies/mL)/ CMV
therapy
DNAemia (duration in
Log2 Baseline AbNEI
Polyfunctional
titer/Peak AbNEI titer *
CD8+ T cells (cells/μL) †
days)
25
1
D+R+
Yes
Yes
1062 / 45
7.85 / 7.85
NA
2
D-R+
Yes
Yes
13165 / 40
7.85/ 7.85
NA
3
D+R+
Yes
Yes
3656 / 89
5.77 / 5.77
0.00 ††
4
D+R+
Yes
Yes
40851 / 73
5.77 / 5.77
5.83 ††
5
D-/R+
Yes
Yes
1108 / 57
3.69 /4.38
0.00 ††
6
D+R+
Yes
Yes
17270 / 150
NA/ NA
0.00 ††
7
D+R+
Yes
Yes
2842 / 36
7.15 / 7.15
1.10
26
8
D-/R+
Yes
Yes
1455 / 142
5.77 / 5.77
0.06
9
D-/R+
Yes
Yes
2039 / 124
5.77 / 5.77
0.23
10
D-/R+
Yes
Yes
4873 / 127
6.46 / 6.46
0.15
11
D+/R-
Yes
Yes
4828 / 30
2.30 / 3.00
0.00
12
D-/R+
Yes
Yes
2094 / 34
5.77 / 6.46
0.00
13
D-R+
Yes
No
272 / 115
6.46 / 6.46
NA
14
D+R+
Yes
No
725 / 64
7.15 / 7.15
NA
15
D+R+
Yes
No
94 / 19
NA / NA
NA
16
D+R+
Yes
No
106 / 23
5.08 / 5.08
2.49 ††
17
D+R+
Yes
No
74 / 44
6.46 / 7.15
1.66 ††
18
D-/R+
Yes
No
277 / 49
3.69 / 3.69
0.00 ††
19
D+R+
Yes
No
144 / 67
7.15 / 7.15
0.00
20
D+R+
Yes
No
53 / 18
7.15 / 7.85
0.94
27
21
D+R+
Yes
No
593 / 36
6.46 / 6.46
0.31
22
D-/R+
Yes
No
60 / 7
7.15 / 7.15
0.02
23
D+R+
Yes
No
53 / 7
4.38 / 4.38
1.49
24
D+R+
Yes
No
907 / 41
5.08 / 5.08
0.00
25
D+R+
No
NA
NA
6.46 / 6.46
NA
26
D+R+
No
NA
NA
5.08 / 5.08
NA
27
D+R+
No
NA
NA
3.00 / 3.00
NA
28
D-/R+
No
NA
NA
1.61 / 5.08
NA
29
D-/R+
No
NA
NA
1.61 / 1.61
NA
30
D+R+
No
NA
NA
3.69 / 4.38
0.79
31
D-/R+
No
NA
NA
7.85 / 7.85
0.74
32
D-/R+
No
NA
NA
4.38 / 4.38
0.09
33
D+R+
No
NA
NA
2.30 / 3.69
0.02
34
D-/R+
No
NA
NA
1.61 / 3.00
0.49
35
D+R+
No
NA
NA
4.38 / 4.38
2.01
36
D+R+
No
NA
NA
3.69 / 4.38
0.00
37
D-/R+
No
NA
NA
1.61 / 3.00
0.00
38
D+R+
No
NA
NA
1.61 / 1.61
0.00
39
D-/R+
No
NA
NA
1.61 / 1.61
0.00
Abbreviations: AbNEI, antibodies neutralizing epitelial infection; CMV, Cytomegalovirus; D, Donor; NA, not applicable; R, Recipient. *
AbNEI titer (inverse) in patients with or without subsequent CMV DNAemia
† Total cell count of polyfunctionals CD8+ T cells (expressing 2 or more of the following markers: IFN-γ/TNF-α, IFN- γ/CD107a) at day +30 after transplant as determined by flow cytometry for intracellular cytokine staining. †† Patients with CMV DNAemia prior the time of immunological monitoring. 546 547 28
548 549 Table 2. Kinetics of CMV DNAemia in patients displaying baseline and peak cytomegalovirus-specific antibodies neutralizing epithelial infection (AbNEI) either below or above 4.7 log2 titer. Baseline AbNEI Parameter
Median initial CMV DNA load in copies/ml (range) Median peak CMV DNA load in copies/ml (range) Median duration of CMV DNAemia in days (range)
log2 titer (inverse)
Peak AbNEI log2 titer P
(inverse)
value* < 4.7
> 4.7
52 (16-
54 (10-
91)
188)
693 (53-
985 (53-
4828)
40851)
39.5 (7-
44.5 (7-
57)
150)
0.76
0.67
0.30
P value*
< 4.7
> 4.7
51 (16-
55 (10-
91)
188)
1108 (53-
907 (53-
17270)
40851)
49 (7-
44 (7-
150)
142)
0.41
0.72
0.89
*Differences between medians were compared using the Mann-Whitney U-test. Twosided exact P values are reported. A P value 0.1% (mean+2 standard deviation-SD- of CMVseronegative controls) after background substraction for IFN-γ, TNF-α, and CD107aexpressing CD8+ T-cell populations. †Differences between medians were compared using the Mann-Whitney U-test. Twosided exact P values are reported. A P value
