Int..L lmmunopharmac., Vol. 14, No. 4, pp, 573-581, 1992. Printed in Great Britain.
0192-0561/92 $5.00 + .130 Pergamon Press Ltd. © 1992 International Society for Immunopharmacology.
POTENTIATION OF HIV ENVELOPE GLYCOPROTEIN A N D OTHER IMMUNOGENS BY ENDOTOXIN (ET) A N D ITS MOLECULAR FRAGMENTS E. KOVATS,* B. SOMLYO,* E. CSANKY,* X. M. SHI,* Y. L. ZHANG,* R. T. COUGHLIN"t and A. NOWOTNY* *University of Pennsylvania, Center for Oral Health Research, Philadelphia, PA 19104; and *Cambridge Biotech, Worcester, MA 01605, U.S.A. (Received 1 January 1991 and in final form 15 November 1991)
-The structural requirements for the immunopotentiating (adjuvant) effect of endotoxin (ET) were investigated. Mild hydrolysis (0.2 N acetic acid at 95 °C) was applied to various ET preparations and the lipid rich (Lipid A) and polysaccharide-rich (PS) preparations obtained were tested as adjuvants on three immunogens: sheep red blood cells (SRBC), L-glutamine : L-lysine : L-alanine containing random synthetic polypeptide (GLA-40), and recombinant HIV viral envelope polypeptide (CBre3). It was found that not only the Lipid A precipitates, but under certain hydrolytic conditions the non-toxic PS preparations were also potent adjuvants. The exact conditions of hydrolysis which led to the isolation of immune adjuvant bacterial products were established. These materials were also tested for endotoxicity (Limulus lysate clotting, chick embryo lethality and local Shwartzman skin reactivity), as well as for TNF generating activities. It was found that TNF generation runs parallel with toxicity of the samples, but it does not follow the adjuvant activity of the isolates. Chemical analysis of the preparations indicated that they did not contain residual ET or Lipid A, however, they did not exclude that deacylated and dephosphorylated skeletal remains of ET are among those components in these preparations which have immunomodulatory activity. Abstract
Dilute acetic acid (Boivin, Mesrobeanu & Mesrobeanu, 1933a) or hydrochloric acid (Westphal & Luderitz, 1954) hydrolyzes endotoxin (ET), precipitates a mixture of lipid-rich breakdown product Lipid A and leaves in the solution the carbohydrate-rich fragments. The earliest studies on such fractions showed that the lipid precipitate contains toxic remnants of the ET while the PS is neither toxic nor immunogenic. Therefore, much attention has been paid to the chemistry as well as to the biological activities of the Lipid A precipitates and, later, to the lipid moiety of ET, from which the Lipid A precipitate was produced by partial hydrolysis. Studies of the polysaccharide-rich (PS) preparations revealed the composition of the polysaccharide moiety of ET, identified the ospecific immunodeterminants (Luderitz, 1970; Westphal & Luderitz, 1960), but did not include the testing of PS preparations in biological assays of ET activities. We observed reduced but still significant levels of biological activities of PS samples including the generation of colony stimulating factor (CSF) 573
and protection against lethal irradiation of mice (Chang, Thompson & Nowotny, 1974; Nowotny, Behling & Chang, 1975). The immunostimulating effect of PS was also detected: in vivo (Behling & Nowotny, 1977a,b, 1980, 1982) and in vitro (Frank, Specter, Nowotny & Friedman, 1977; Friedman, Butler, Specter & Nowotny, 1988). Independent confirmation of these findings came from several laboratories. Apte, Hertogs & Pluznik (1977) reported that their PS preparation was approximately half as active as ET in the CSF test. The liberation of IL-1 by another PS preparation was reported by Haeffner-Cavaillon, Cavaillon & Szabo (1982, 1984). Raichvarg, Brossard & Agneray (1980) prepared a PS fraction using hydrolysis with ion exchanger resins and Urbaschek (1980) isolated a polysaccharide-rich preparation from whole bacterial cells. These were also active in various assays such as the enhancement of phagocytosis, induction of CSF and resistance to lethal irradiation. As new information in this publication, we will describe the optimal hydrolytic conditions for the
E. KovA-rSet al.
574
liberation of an adjuvant active (ADA) fraction. The endotoxicity of the isolated samples was tested in rabbit local Shwartzman reactivity and chick embryo lethality. The fact of TNF liberating and Limulus lysate coagulating structural subunits during such acidic hydrolysis was also included in these studies.
EXPERIMENTAL PROCEDURES
Mice. CD-1 (ICR) mice, 5 - 6 weeks old, 2 0 - 2 4 g body weight, were purchased from Charles River Co. (Wilmington, MA) for the measurement of immune responses. Endotoxin (ET). Serratia marcescens ATCC No. 13477 was extracted with 5°70 trichloracetic acid (TCA) according to a modified procedure of Boivin, Mesrobeanu & Mesrobeanu (1933b) (Nowotny, 1979). ET was purified by precipitation with ethanol containing 0.2070 MgC12 and by sedimentation at 110,000 g for 3 h. Salmonella minnesota S 1114 was extracted by the 45% phenol method of Westphal & Luderitz (1954) and purified as above. PS and Lipid A. A solution of the ET preparations, 2 mg/ml, was prepared and hydrolyzed in 0.2 N acetic acid at 95°C (_+ 0.1°C). The samples were centrifuged to sediment the Lipid A (10,000g, 60min). The supernatants were lyophilized without dialysis. The Lipid A precipitate was resuspended in distilled water and also lyophilized. All preparations were stored in vacuum desiccators. Immunogens and immunization. Sheep red blood cells (SRBC) were freshly drawn and provided in Alsevier solution by the School of Veterinary Medicine of the University of Pennsylvania. CD-1 mice received 107 washed SRBC in 0.25 ml volume i.p. The immune response to SRBC was determined 4 days later using the Jerne plaque assay (Nowotny, 1979). The recombinant HIV envelope protein preparation (named CBre3) containing one-third of the carboxy terminus of gpl20 and a half of the N terminus of gp 41 (Thorn et al., 1987) was provided by Cambridge Biotech (Worcester, MA). CBre3, 5 ~g, was injected into CD-I mice i.p. with or without ET and its derivatives. The adjuvant test samples were 10/ag/mouse given i.p. at the same time. Ten mice were used in each group. The antibody levels of pooled sera were measured 7 days later by an indirect ELISA. In brief, 96-well microtiter plates (Becton - Dickinson, Lincoln Park, N J) were coated with CBre3. The blocking solution contained 4°70 bovine serum albumin at pH 5. The
washed plates received the mouse sera at four-fold dilutions, incubated for 2 h, washed, and the second antibody (alkaline phosphatase labeled goat antimouse IgG) added. Two hours incubation at 37°C was followed by washing the plates and by pipetting the substrate (p-nitrophenyl phosphate) into the wells. The developed color was read 15, 30 and 45 min later at 405 nm, using the Titertek ~ Multiscan plate reader (Flow Laboratories, McLean, VA). The third immunogen was a synthetic random polypeptide of L-glutamic acid-L-lysine-L-alanine (GLA-40) in a 36 : 24 : 40 ratio. This material was kindly provided by Dr Paul Maurer (Jefferson Medical University, Philadelphia, PA). This synthetic polypeptide 10 t~g, was injected i.p. into CD-1 mice with or without ET derivatives. Seven days later the anti-GLA titer of the sera was determined by ELISA using a procedure similar to the one described above. Determination o f the adjuvant activity (ADA ). The immunogens were injected alone or with 10/ag of the adjuvant sample as described above. In the case of CBre3 or GLA-40 injections, the mice were sacrificed 7 days later and their sera collected for ELISA. The ADA was expressed as Stimulation Index (SI) where SI = ELISA signal (OD reading) from adjuvant injected immunized mice minus ELISA signal from non-immune mice divided by ELISA signal from immunized mice minus ELISA signal from non-immune mice. The SI for SRBC: the OD readings in above equation were replaced by the number of plaques per 10 6 spleen cells. Polyclonal B-cells activation (PBA ). The mice received 10/ag of the adjuvant preparations without CBre3 and the CBre3-reactive immunoglobulin titers were measured 7 days later. Assay f o r the measurement o f total IgG release. Microtiter plates were coated with goat anti-mouse IgG (Sigma Co., St. Louis, MO, cat. No. 8642) using a carb0nate-bicarbonate buffer (0.05 M, pH 9.6). The washed plates received 50/al of serially diluted test sera using a T w e e n - p h o s p h a t e buffered saline diluent. After incubations and washings, the wells received goat anti-mouse IgG labeled with alkaline phosphatase (Sigma Co, St. Louis, MO, cat. No. 5153), and the plates were incubated for 90 min and washed before the p-nitrophenyl phosphate substrate was added to each well. The developed color was read at 405 nm. TNF assay. RAW 264.7 cells were maintained in DMEM medium and exposed to ET or to ET derivatives. The supernatant was added to L929 target cells in various dilutions. Cytotoxicity was
Potentiation of HIV Envelope Glycoproteins measured by staining the surviving and still adhering L929 cells with crystal violet according to the method of Ruff & Gifford (1981). Determination o f endotoxicity. Chick embryo lethality (Smith & Thomas, 1956) and Shwartzman skin reactivity in rabbits were used to estimate the in vivo toxicity of the isolates according to the methods described elsewhere (Nowotny, 1979). The chromogenic LAL kit of Whittaker M.A. Bioproducts (Walkersville, MD) was used to quantitate ET content according to the instructions of the manufacturer. The ET content was expressed as ET units. Chemical analysis. The percent total reducing carbohydrate, hexosamine, carboxylic acid and amino acid content of the samples was carried out as described before (Nowotny, 1979). Qualitative analysis of the carboxylic acids was carried out by GLC of their methyl esters (Nowotny, 1979). For g a s - l i q u i d chromatography of fatty acids, 1 mg samples were transesterified using BF3 in anhydrous methanol as described elsewhere (Nowotny, 1979). The hexane extract of the fatty acid methyl esters was used to quantitate the ester content by the hydroxylamine procedure. Five microliter atiquots were injected into a Hewlett Packard (Palo Alto, CA) 5830-A GLC for the identification of the methyl esters. Chromatographic conditions: 10070 SP-2330 on 100-120 mesh Supelcoport column (Supelco, Inc., Bellefonte, PA). The temperature was 100- 220°C, with an 8°C/min gradient. Dry-weight determination was done by pipetting 100~1 aliquots onto a pre-weighed 20 x 20 mm aluminum foil piece and drying under an infrared lamp oven and weighed using an ultra microbalance. If the sample contained salts or buffers, the extracts were dialyzed exhaustively prior to dry-weight content determination.
RESULTS
Effect o f acidic hydrolysis on the adjuvant activity (ADA)
The ADA of hydrolyzates of S. marcescens ATCC No. 13477 ET with 0.2 N acetic acid at 95°C is shown in Fig. 1. The immunogen in these experiments was freshly drawn sheep red blood cell (SRBC) suspension. The value 1.0 is the immune response elicited by SRBC alone. It is evident from this figure that short hydrolysis produces a potent Lipid A precipitate and a PS which is inert. Longer hydrolysis apparently released adjuvant active
575
SERRATIA MARCESCENS A'B'CC 13477
,.o
I I
I
I 3.0
I PS
.~_ --~ ==
2.0
1.0
• SRBC only
I
I
0'
I
I
I
I
30' 60' 90' 120'
240'
Time of hydrolysis (min)
Fig. 1. Immune response to SRBC in mice. Effect of continuous mild hydrolysis on the immunomodulating
activity of ET from S. marcescens ATCC 13477.
SERRATIA
MARCESCENS
ATTCC
13477
4.0
\LIPIDA 2.0-
1.0 .
.
0'
I
.
I
.
I
I
CBre3 only
I
30' 60' 90' 120' 240' Time of hydrolysis (min)
Fig. 2. Immune response to recombinant HIV glycoprotein CBre3 in mice. Effect of continuous mild hydrolysis on the immunomodulating activity of ET from S. marcescens ATCC 13477. Values less than 1.0 stimulation index mean immunosuppression. components into the supernatant, and this release reaches its maximum around 120 min. Continued hydrolysis destroys the adjuvant activity of the PS components. The same S. marcescens ATCC 13477 ET and its split products were also tested for adjuvant activity
576
E. KOVATS el al. Serratla
marcescens
ATCC
13477
SAMONELLA
4.0-
3.0-
'° 1
/
MINNESOTA
1114
-. j,
3.0-
._~ E
N 20
2.0-
2
1.0-
GLA 40 only
1.0
CBre3 only
I
I
0' Time of hydrolysis (rain)
Fig. 3. Immune response to random synthetic polypeptide GLA-40 in mice. Effect of continuous mild hydrolysis on the immunomodulating activity of ET from S. marceseens ATCC 13477. in c o m b i n a t i o n with CBre3 a n d G L A - 4 0 polypeptide i m m u n o g e n s . T h e results were similar to those observed using S R B C as the i m m u n o g e n . S h o r t hydrolysis p r o d u c e d a p o t e n t Lipid A a d j u v a n t a n d an inert, or in some cases, an a p p a r e n t l y i m m u n o s u p p r e s s i v e PS. If the hydrolysis was c o n t i n u e d for 60 a n d 120 m i n the PS b e c a m e active but f u r t h e r hydrolysis destroyed the a d j u v a n t activity o f the PS. In some cases it restored the i m m u n e e n h a n c i n g p o w e r o f the Lipid A. Figures 2 a n d 3 show these findings. If S. m i n n e s o t a S 1114 E T was subjected to identical hydrolysis, a n d samples were t a k e n for A D A d e t e r m i n a t i o n d u r i n g the time course o f the e x p e r i m e n t , the p a t t e r n o f the changes o f A D A was s o m e w h a t different f r o m those s h o w n in Figs 1, 2 a n d 3 for S. m a r c e s c e n s A T C C 13477. T h e Lipid A p r e p a r a t i o n s were m o r e p o t e n t t h a n the PS samples d u r i n g the first 2 h o f hydrolysis, but the PS gained A D A if the hydrolysis was c o n t i n u e d for 4 h. T h e Lipid A also showed increasing A D A d u r i n g the same time. Figure 4 presents the SI values for CBre3 as a n i m m u n o g e n .
I
I
~
I
30' 60' 90' 120' 240' Time of hydrolysis (rain)
Fig. 4. Immune response to recombinant HIV glycoprotein CBre3 in mice. Effect of continuous mild hydrolysis on the immunomodulating activity of ET from S. minnesota 1114.
Table 1. Polyclonal B-cell activation by S. marcescens ATCC 13477 preparations Preparations*
Increase of anti-CBre3 immunoglobulin levels ~
ET
2.7
30 60 120 240
PS min min min min
95°C 95°C 95°C 95°C
1.3 1.2 1.2 0.9
30 60 120 240
Lipid A min 95°C min 95°C min 95°C min 95°C
2.4 2.1 1.6 1.2
*10 Mg were injected i.p. into CD-I mice. *The anti-CBre3 level of saline and of preparation injected mouse sera were determined by ELISA. The numbers ion this column indicate the x fold increase in the anti-CBre3 globulin levels.
Polyclonal B-cell activations
M o s t ET p r e p a r a t i o n s tested so far showed some activity in releasing CBre3-reactive i m m u n o g l o b u l i n s in vivo. T h e Lipid A samples were also active; in a
few cases as active as the ET samples but the PS showed insignificant P B A i n d u c t i o n (Tables 1 a n d 2).
Potentiation of HIV Envelope Glycoproteins Table 2. Polyclonal B-cell activation by S. minnesota 1114 and preparations Preparations*
Increase of anti-CBre3 immunoglobulin levels~
ET Sl114 R595
2.2 2.7
PS* min 95°C min 95°C min 95°C min 95°C
1.0 0.6 1.6 0.2
Lipid A* 30 min 95°C 60 min 95°C 120 min 95°C 240 min 95°C
2.8 2.3 0.4 NT
30 60 120 240
*10 tag were injected i.p. into CD-1 mice. +The anti-CBre3 level of saline and of preparation injected mouse sera were determined by ELISA. The numbers in this column indicate the x fold increase in the anti-CBre3 immunoglobulin levels. *Preparations obtained from S. minnesota S1114.
Table 3. Release of total serum IgG by S. minnesota 1114 preparations Preparations*
Increase of total serum IgG levelst
Saline
1.0
ET S1114 R595
10.5 2.9
PS* 30 min 60 min 120 min 240 min
8.6 2.7 7.1 0.9
Lipid A* 30 min 60 min 120 min 240 min
10.1 8.0 1.8 NT
*10 ~g were injected i.p. into CD-1 mice. tThe IgG level of saline and of preparations injected mouse sera were determined by ELISA. The numbers in this column indicate the x fold increase in the immunoglobulin levels. *These preparations were obtained from S. minnesota S1114.
577
Release o f serum IgG T h e t o t a l IgG level in the sera of the injected mice was o f t e n m a n y fold increased as c o m p a r e d with the saline injected CD-1 mice. The m o s t active were S. minnesota S 1114 ET, the 30 a n d 120 m i n PS, the 30 a n d 60 m i n Lipid A f r o m the same s t r a i n ' s E T (Table 3). Some o f these p o t e n t a d j u v a n t , P B A active or IgG level e n h a n c i n g p r e p a r a t i o n s were n o longer endotoxic (Table 4).
T N F and L A L activities o f ET, Lipid A and PS T a b l e 4 presents the results o f T N F a n d L A L activity changes d u r i n g acidic hydrolysis. G o o d c o r r e l a t i o n has been f o u n d between the L A L positivity a n d T N F inducing capacity of the fractions. Neither the T N F n o r L A L activity correlated with the immune responsiveness e n h a n c i n g potency o f the PS samples.
Endotoxicity o f the hydrolyzed preparations Table 4 also shows the changes in chick e m b r y o lethality a n d in the local S h w a r t z m a n reaction d u r i n g hydrolysis with 0.2 N acetic acid. T h e endotoxicity o f the samples rapidly diminishes d u r i n g such treatment.
Chemical analysis T a b l e 5 shows the c o m p o s i t i o n o f S. marcescens A T C C 13477 ET, a n d PS extracts. The c o m p o n e n t s o f the E T are similar to those reported for E T p r e p a r a t i o n s isolated by the T C A p r o c e d u r e o f Boivin a n d associates. The analysis o f the PS o b t a i n e d by 0.2 N acetic acid cleavage at 95°C in 120 m i n shows two m a j o r changes as c o m p a r e d with the starting material (ET) : the e l i m i n a t i o n o f fatty acids a n d p h o s p h o r u s . A p p r o x i m a t e l y 80°7o o f the fatty acid loss occurs in the first 30 min o f hydrolysis (data not shown).
G L C o f fatty acids The fatty acid c o n t e n t of S. marcescens E T samples varied between 16 a n d 17°70. PS p r e p a r a t i o n s o b t a i n e d after 120 or 240 rain hydrolyzes at 95°C in 0.2 N acetic acid c o n t a i n only trace a m o u n t s o f fatty acid ( < 0 . 5 % ) . Figure 5a shows the fatty acid c o n t e n t o f S. mareescens E T a n d Fig. 5b shows t h a t o f a 120 m i n PS sample, carried o u t u n d e r identical conditions.
578
E. KOVATS et al.
Table 4. Changes in the TNF generating and endotoxic activity of the preparations during acidic hydrolysis. Preparations
TNF*
LAL +
SHWM*
LD50 ~
100 100 --
1.8 x 106 3.4 X 1 0 7 6.9 × 107
4.5 3.8 4.1
0.015 0.022 0.080
ET S. marcescens S. minnesota Sl114 S. minnesota R595
PS S. marcescens
30 60 120 240
min min min min
Neg. ----
2.8 × 10~ 0.3 × 10' Neg. Neg.
1.4 0.3 Neg. Neg.
0.92 5.00 10.00 10.00
30 60 120 240
min min min min
-----
3.5 × 105 3.3 × l0 s Neg. Neg.
2.5 1.3 0.3 Neg.
1.23 5.00 10.00 10.00
30 60 120 240
min min min min
52 32 24 12
9.6 4.7 2.8 0.2
× × × ×
105 105 105 103
3.2 2.2 2.9 2.5
0.09 0.03 0.10 0.25
30 60 120 240
min min min min
-----
5.5 4.2 3.3 7.4
× × >( ×
107 107
2.9 2.7 2.0 0.8
0.02 0.02 0.15 0.09
S. minnesota 1114
Lipid A S. marcescens
S. minnesota SI 114
10 7
105
*TNF units/ml are expressed according to the procedure of Ruff & Gifford (1981). +Limulus lysate chromogenic assay results expressed as ET units per/ag preparations.
*Local Shwartzman reaction. Numbers show size of hemorrhagic area in square cm. ~Chick embryo lethality LDs0, in ~g. (--) Not tested. (Neg.) No reaction. DISCUSSION To study the molecular aspects o f i m m u n e p o t e n t i a t i o n by ET and its analogs our a p p r o a c h is the identification o f the minimal structural requirements in the macromolecule o f ET or its c o n t a m i n a n t s , which are able to induce i m m u n e enhancement. The use o f fragments for the identification o f A D A active structural elements o f ET might need explanation, particularly these days when synthetic c o m p o u n d s with c o m p a r a b l e endotoxicity are available. First o f all, these synthetic c o m p o u n d s could be synthesized only after the structure o f the lipid moiety had been elucidated by a n u m b e r o f earlier investigators. The m o n o p h o s p h o r y l lipid A ( M P L ) p r e p a r a t i o n s are p o t e n t adjuvants, but they
are complex mixtures o f n u m e r o u s split p r o d u c t s and it has not been firmly established which one o f these or which c o m b i n a t i o n o f these is required for A D A . As will be discussed below, our adjuvant active PS preparations do not show any resemblance to Lipid A c o m p o n e n t s or M P L . They represent a new structural entity o f ET and A D A . The aim o f our studies, therefore, is first the exact identification o f the structure o f the i m m u n o p o t e n t i a t i n g PS c o m p o n e n t , and if we succeed with these efforts, the next step will be a t t e m p t s to synthesize such components. As far as the i m m u n o g e n s are concerned, we selected three: the convenient SRBC, a synthetic r a n d o m polypeptide GLA-40 and a r e c o m b i n a n t HIV-1 envelope protein, CBre3. We carried out preliminary experiments to establish the dose o f the
Potentiation of HIV Envelope Glycoproteins Table 5. Chemical analysis of ET and PS preparations Constiuents %
ET*
PSi
Amino acids* Amino sugars ~ Reducing carbohydrates H KdoI Fatty acids** Phosphorus t* Nucleic acids**
8.14 14.32
3.21 18.10
31.8 1.46 16.06 1.02 2.30
46.96 1.57 0.21 0.05 3.75
*TCA extracted and purified ET from S. marcescens ATCC 13477. *PS from above ET obtained by hydrolysis with 0.2 N acidic acid at 95°C for 120 min. *Expressed a % glycine, bassed on primary amine determination and corrected for hexosamines. ~Expressed as % D-glucosamine. "Expressed as % D-glucose. I°70 2-keto-3-deoxy octulosonic acid. **Expressed as % palmitic acid. t'Total % phosphorus content. t'Total % nucleic acid content.
11.83 14.15 .9:, I /16.12
B
li 21.10
i
l
21.10
1.1_ Fig. 5. GLC of fatty acids. (A) Fatty acid methylesters in TCA extracted and purified ET from S. marcescens ATCC 13477. (B) Same analysis carried out on a 120 rain 95°C hydrolyzed PS preparation. immunogens so that a low but definite immune response would be elicited without adjuvants. In the experiment reported here, the adjuvants were given at the same time as the immunogen but by separate injections, and both were injected i.p. These
579
experimental conditions leave a number of other variables to be tested, such as variations of the ratio between the dose of immunogen and immune stimulator and the time interval between the injection of the adjuvant and the immunogen. Since the adjuvant potentials of the PS were reported earlier by us, it may be necessary to underscore the information in this publication which is unquestionably new. The first is that potent immune enhancers are among the partial hydrolytic products of ET, and that these do not elicit any of the effects of ET which are considered to be undesirable. With the exception of 30 min PS preparations, these adjuvant active isolates were negative even in the lymphocyte proliferation (mitogenicity) test (Frank et al., 1977; Friedman et al., 1988). Our earlier claim regarding the immunopotentiating effect of PS preparations has been questioned by the more traditional investigators of structure and function relationship in endotoxins. We believe that these new data give further substantiation to our claim by showing that several PS preparations have a definite immune adjuvant effect, tested on three different immunogens. The second point for discussion deals with the possible origin of the molecular fragments with adjuvant activity. The rather obvious question is whether they are split products of the lipid moiety or of the polysaccharide regions or non-endotoxic contaminants of the TCA extracted ET. Chemical analysis has been carried out so far only on a few samples, which showed the highest ADA. These limited data show rather convincingly that the PS do not contain residual ET. The fact that one of the most active PS samples does not contain more than trace amounts of fatty acid and phosphorus seem to indicate that these compounds are not contaminated with Lipid A either. It remains to be investigated whether these preparations are split products of the core region and whether they may contain deacylated fragments of the glucosamine backbone of the lipid moiety. Our chemical analysis also rules out the possibility that the PS preparations would contain " m o n o phosphoryl E T " (MPL) molecules (Qureshi, Takayama & Ribi, 1985). MPL is obtained by very mild acidic hydrolysis of mutant ET (DPL) and it is claimed that this treatment cleaved only the glycosidically bound phosphate group leaving the fatty acids intact rendering the ET non-toxic but still an adjuvant (Johnson & Tomai, 1990). We detected almost no phosphorus or fatty acid in PS preparations.
580
E. KOVATSet al.
We think that another noteworthy observation was the generation of CBre3-reactive immunoglobulins into the sera of mice which received adjuvants but no antigen. The immunochemical characterization of these immunoglobulins and particularly their specificity for the H I V envelope remain to be elucidated. Differences were found between S. marcescens A T C C 13477 and S. m i n n e s o t a S 1114 ET derived PS and Lipid A preparations tested as adjuvants. The unique patterns o f Figs 1, 2 and 3 show interesting changes in the adjuvant activity of the PS and Lipid A samples during acidic hydrolysis. Similar, but not identical, changes could be observed during the hydrolysis of S. m i n n e s o t a ET. This nonidentity might be attributed to either structural differences between the two ET preparations, or to the presence of components in the S. marcescens ET which were not present in the S. m i n n e s o t a sample. Since we found significant differences among the A D A of a few Gram-negative strains ET or PS, one should expect that similar or greater differences will be found if one chooses other Gram-negative strains for similar studies. Whether our findings are applicable to all G r a m negative bacteria and their products remains to be investigated. These experiments we report here were carried out by giving one i m m u n o g e n and one adjuvant injection
and the mice were sacrificed 7 days later. A number of further variations are carried out at the present time in our laboratory to optimize the A D A of our samples. These include repeated injections of CBre3 a n d / o r PS, variations of the time intervals between these two and measurements of the duration of the elevated anti-CBre3 levels in the sera of the mice. The possible existence of immunosuppressive structural elements in the ET samples we studied needs rigorous substantiation. It was shown unequivocally that the PS obtained after short (30 min) hydrolysis is either inert as an adjuvant or in some systems immunosuppressive as shown in Figs 1 and 2. Under the same conditions the 30 min Lipid A precipitate was active but became inert or slightly suppressive after 60 min. We decline to speculate at this time about the possible explanations of the drastic changes in A D A of the PS or Lipid A samples occurring during continued acidic hydrolysis but we pursue this observation by attempting to isolate immune suppressive split products during such hydrolytic processes. Acknowledgements - - This work has been supported by
NIH grants AI25348, CA24628 and by a donation received from Cell Technology, Inc., Boulder, CO. The authors gratefully acknowledge the provision of CBre3 recombinant HIV glycoprotein by Cambridge Bioscience, Worcester, MA.
REFERENCES
APTE, R. M., HERTOGS, C. F. & PLUZNIK, D. H. (1977). Regulation of lipopolysaccharide-induced granulopoiesis and macrophage formation by spleen cells. I. Relationship between colony-stimulating factor release and lymphocyte activation in vitro. J. Immun., 118, 1435 - 1440. BEHEING, U. H. & NOWOTNY, A. (1977a). Immune adjuvancy of lipopolysaccharide and a nontoxic hydrolytic product demonstrating oscillating effect with time. J. Immun., 118, 1905 - 1907. BEHLING, U. H. & NOWOTNY,A. (1977b). Long-term adjuvant effect of bacterial endotoxin in prevention and restoration of radiation-caused immuno-suppression. Proc. Soc. exp. Biol. Med., 157, 348- 353. BEHLING, U. H. & NOWOTNY,A. (1980). Cyclic changes of positive and negative effects of single endotoxin injections. In Bacterial Endotoxin and Host Responses (ed. Agarwal, M. K.), pp. 11-26. Elsevier/North Holland Biomedical Press, Amsterdam. BEHLING, U. H. & NOWOTNY,A. (1982). Bacterial endotoxins as modulators of specific and non-specific immunity. In Regulatory Implication o f Oscillatory Dynamics in the Immune Response (eds Hiernaux, J. and DeLisi, C.), Vol. 2, pp. 127- 140. CRC Press, Boca Raton, FL. BOIVIN, A., MESROBEANU,J. & MESROBEANU,L. (1933a). Technique pour la preparation des polysaccharides microbiens specifiques. Compt. Rend. Soc. Biol., 113, 490-492. BOIV1N, A., MESROBEANU,J. & MESROBEANU,L. (1933b). Extraction d'un complexe toxique et antigenique a partir du bacille d'Aertrycke. Comp. Rend. Soc. Biol., 114, 307-310. CHANG, H., THOMPSON, J. J. t~£ NOWOTNY, A. (1974). Release of colony stimulating factor (CSF) by non-endotoxic breakdown products of bacterial lipopolysaccharides. Immun. Commun., 3, 401-409. FRANK, S., SPECTER, S., NOWOTNY, A. • FRIEDMAN, H. (1977). Immunocyte stimulation in vitro by nontoxic bacterial lipopolysaccharide derivatives. J. Immun., 119, 855-869.
Potentiation of HIV Envelope Glycoproteins
581
FRIEDMAN, H., BUTLER, R. C., SPECTER, S. & NOWOTNY, A. (1988). Nontoxic endotoxin polysaccharide induces soluble mediators which potentiate antibody production by murine retrovirus suppressed splenocytes. Int. J. Immunopharmac., 10, 283 - 292. HAEEFNER-CAVAILLON, N., CAVAILLON, J. M. & SZABO, L. (1982). Macrophage-dependent polyclonal activation of splenocytes by Bordetella pertussis endotoxin and its isolated polysaccharide and lipid A regions. Cell. Immun., 74, 1-13. HAEFFNER-CAVAILLON,N., CAVAILLON,J. M. & SZABO, L. (1984). Interleukin 1 secretion by human monocytes stimulated by the isolated polysaccharide region of Bordetella pertussis endotoxin. Molec. Immun., 21, 389- 395. JOHNSON, A. G. & TOMAI, M. A. (1990). A study of the cellular and molecular mediators of the adjuvant action of a nontoxic monophosphoryl Lipid A. Adv. exp. Med. Biol., 256, 567-579. LUDERITZ, O. (1970). Recent results on the biochemistry of the cell wall lipopolysaccharides of Salmonella bacteria. Angew. Chem., 9, 649-663. NOWOTNY, A. (1979). Basic Exercises in Immunochemistry, 2nd Edn. Springer, New York. NOWOTNY, A., BEHLING, U. H. & CHANG,H. L. (1975). Relation of structure to function in bacterial endotoxins. VIII. Biological activities in a polysaccharide-rich fraction. J. Immun., 115, 199-203. QURESHI, N., TAKAYAMA,K. & RIBI, E. (1982). Purification and structural determination of nontoxic Lipid A obtained from the lipopolysaccharide of Salmonella typhymurium. J. biol. Chem., 257, 11808- 11815. RAICHVARG, D., BROSSARD, C. & AGNERAY, J. (1980). Preparation of a nontoxic and immunogenic polysaccharide fraction from a Haemophilus influenzae phenol-water extract. Infect. Immun., 29, 171 -174. RUFF, M. R. & GIFrORD, G. E. (1981). Rabbit tumor necrosis factor: mechanism of action. Infect. Immun., 3 1 , 3 8 0 - 385. SMITH, R. T. & THOMAS, L. (1956). The lethal effect of endotoxin on the chick embryo. J. exp. Med., 104, 217-231. THORN, R. M., BELTZ, G. A., HUNG, C., FALLIS, B. F., WINKLE, S., CHENG, K. & MARCIANI,D. J. (1987). Enzyme immunoassay using a novel recombinant polypeptide to detect human immunodeficiency virus env antibody. J. cfin. Microbiol., 25, 1207- 1212. URBASCHEK,R. (1980). Effects of bacterial products of granulopoiesis. Adv. exp. Med. Biol., 121b, 5 1 - 64. WESTPHAL, O. & LUDERITZ,O. (1954). Chemische Erforschung yon Lipopolysacchariden Gram-negativer Bakterien. Angew. Chem., 66, 407-417. WESTPHAL,O. & EUDERITZ,O. (1960). 3.6-didesoxy-hexosen-Chemie und Biologie. Angew. Chem., 72, 881- 891.
0192-0561/92 $5.00 + .130 Pergamon Press Ltd. © 1992 International Society for Immunopharmacology.
POTENTIATION OF HIV ENVELOPE GLYCOPROTEIN A N D OTHER IMMUNOGENS BY ENDOTOXIN (ET) A N D ITS MOLECULAR FRAGMENTS E. KOVATS,* B. SOMLYO,* E. CSANKY,* X. M. SHI,* Y. L. ZHANG,* R. T. COUGHLIN"t and A. NOWOTNY* *University of Pennsylvania, Center for Oral Health Research, Philadelphia, PA 19104; and *Cambridge Biotech, Worcester, MA 01605, U.S.A. (Received 1 January 1991 and in final form 15 November 1991)
-The structural requirements for the immunopotentiating (adjuvant) effect of endotoxin (ET) were investigated. Mild hydrolysis (0.2 N acetic acid at 95 °C) was applied to various ET preparations and the lipid rich (Lipid A) and polysaccharide-rich (PS) preparations obtained were tested as adjuvants on three immunogens: sheep red blood cells (SRBC), L-glutamine : L-lysine : L-alanine containing random synthetic polypeptide (GLA-40), and recombinant HIV viral envelope polypeptide (CBre3). It was found that not only the Lipid A precipitates, but under certain hydrolytic conditions the non-toxic PS preparations were also potent adjuvants. The exact conditions of hydrolysis which led to the isolation of immune adjuvant bacterial products were established. These materials were also tested for endotoxicity (Limulus lysate clotting, chick embryo lethality and local Shwartzman skin reactivity), as well as for TNF generating activities. It was found that TNF generation runs parallel with toxicity of the samples, but it does not follow the adjuvant activity of the isolates. Chemical analysis of the preparations indicated that they did not contain residual ET or Lipid A, however, they did not exclude that deacylated and dephosphorylated skeletal remains of ET are among those components in these preparations which have immunomodulatory activity. Abstract
Dilute acetic acid (Boivin, Mesrobeanu & Mesrobeanu, 1933a) or hydrochloric acid (Westphal & Luderitz, 1954) hydrolyzes endotoxin (ET), precipitates a mixture of lipid-rich breakdown product Lipid A and leaves in the solution the carbohydrate-rich fragments. The earliest studies on such fractions showed that the lipid precipitate contains toxic remnants of the ET while the PS is neither toxic nor immunogenic. Therefore, much attention has been paid to the chemistry as well as to the biological activities of the Lipid A precipitates and, later, to the lipid moiety of ET, from which the Lipid A precipitate was produced by partial hydrolysis. Studies of the polysaccharide-rich (PS) preparations revealed the composition of the polysaccharide moiety of ET, identified the ospecific immunodeterminants (Luderitz, 1970; Westphal & Luderitz, 1960), but did not include the testing of PS preparations in biological assays of ET activities. We observed reduced but still significant levels of biological activities of PS samples including the generation of colony stimulating factor (CSF) 573
and protection against lethal irradiation of mice (Chang, Thompson & Nowotny, 1974; Nowotny, Behling & Chang, 1975). The immunostimulating effect of PS was also detected: in vivo (Behling & Nowotny, 1977a,b, 1980, 1982) and in vitro (Frank, Specter, Nowotny & Friedman, 1977; Friedman, Butler, Specter & Nowotny, 1988). Independent confirmation of these findings came from several laboratories. Apte, Hertogs & Pluznik (1977) reported that their PS preparation was approximately half as active as ET in the CSF test. The liberation of IL-1 by another PS preparation was reported by Haeffner-Cavaillon, Cavaillon & Szabo (1982, 1984). Raichvarg, Brossard & Agneray (1980) prepared a PS fraction using hydrolysis with ion exchanger resins and Urbaschek (1980) isolated a polysaccharide-rich preparation from whole bacterial cells. These were also active in various assays such as the enhancement of phagocytosis, induction of CSF and resistance to lethal irradiation. As new information in this publication, we will describe the optimal hydrolytic conditions for the
E. KovA-rSet al.
574
liberation of an adjuvant active (ADA) fraction. The endotoxicity of the isolated samples was tested in rabbit local Shwartzman reactivity and chick embryo lethality. The fact of TNF liberating and Limulus lysate coagulating structural subunits during such acidic hydrolysis was also included in these studies.
EXPERIMENTAL PROCEDURES
Mice. CD-1 (ICR) mice, 5 - 6 weeks old, 2 0 - 2 4 g body weight, were purchased from Charles River Co. (Wilmington, MA) for the measurement of immune responses. Endotoxin (ET). Serratia marcescens ATCC No. 13477 was extracted with 5°70 trichloracetic acid (TCA) according to a modified procedure of Boivin, Mesrobeanu & Mesrobeanu (1933b) (Nowotny, 1979). ET was purified by precipitation with ethanol containing 0.2070 MgC12 and by sedimentation at 110,000 g for 3 h. Salmonella minnesota S 1114 was extracted by the 45% phenol method of Westphal & Luderitz (1954) and purified as above. PS and Lipid A. A solution of the ET preparations, 2 mg/ml, was prepared and hydrolyzed in 0.2 N acetic acid at 95°C (_+ 0.1°C). The samples were centrifuged to sediment the Lipid A (10,000g, 60min). The supernatants were lyophilized without dialysis. The Lipid A precipitate was resuspended in distilled water and also lyophilized. All preparations were stored in vacuum desiccators. Immunogens and immunization. Sheep red blood cells (SRBC) were freshly drawn and provided in Alsevier solution by the School of Veterinary Medicine of the University of Pennsylvania. CD-1 mice received 107 washed SRBC in 0.25 ml volume i.p. The immune response to SRBC was determined 4 days later using the Jerne plaque assay (Nowotny, 1979). The recombinant HIV envelope protein preparation (named CBre3) containing one-third of the carboxy terminus of gpl20 and a half of the N terminus of gp 41 (Thorn et al., 1987) was provided by Cambridge Biotech (Worcester, MA). CBre3, 5 ~g, was injected into CD-I mice i.p. with or without ET and its derivatives. The adjuvant test samples were 10/ag/mouse given i.p. at the same time. Ten mice were used in each group. The antibody levels of pooled sera were measured 7 days later by an indirect ELISA. In brief, 96-well microtiter plates (Becton - Dickinson, Lincoln Park, N J) were coated with CBre3. The blocking solution contained 4°70 bovine serum albumin at pH 5. The
washed plates received the mouse sera at four-fold dilutions, incubated for 2 h, washed, and the second antibody (alkaline phosphatase labeled goat antimouse IgG) added. Two hours incubation at 37°C was followed by washing the plates and by pipetting the substrate (p-nitrophenyl phosphate) into the wells. The developed color was read 15, 30 and 45 min later at 405 nm, using the Titertek ~ Multiscan plate reader (Flow Laboratories, McLean, VA). The third immunogen was a synthetic random polypeptide of L-glutamic acid-L-lysine-L-alanine (GLA-40) in a 36 : 24 : 40 ratio. This material was kindly provided by Dr Paul Maurer (Jefferson Medical University, Philadelphia, PA). This synthetic polypeptide 10 t~g, was injected i.p. into CD-1 mice with or without ET derivatives. Seven days later the anti-GLA titer of the sera was determined by ELISA using a procedure similar to the one described above. Determination o f the adjuvant activity (ADA ). The immunogens were injected alone or with 10/ag of the adjuvant sample as described above. In the case of CBre3 or GLA-40 injections, the mice were sacrificed 7 days later and their sera collected for ELISA. The ADA was expressed as Stimulation Index (SI) where SI = ELISA signal (OD reading) from adjuvant injected immunized mice minus ELISA signal from non-immune mice divided by ELISA signal from immunized mice minus ELISA signal from non-immune mice. The SI for SRBC: the OD readings in above equation were replaced by the number of plaques per 10 6 spleen cells. Polyclonal B-cells activation (PBA ). The mice received 10/ag of the adjuvant preparations without CBre3 and the CBre3-reactive immunoglobulin titers were measured 7 days later. Assay f o r the measurement o f total IgG release. Microtiter plates were coated with goat anti-mouse IgG (Sigma Co., St. Louis, MO, cat. No. 8642) using a carb0nate-bicarbonate buffer (0.05 M, pH 9.6). The washed plates received 50/al of serially diluted test sera using a T w e e n - p h o s p h a t e buffered saline diluent. After incubations and washings, the wells received goat anti-mouse IgG labeled with alkaline phosphatase (Sigma Co, St. Louis, MO, cat. No. 5153), and the plates were incubated for 90 min and washed before the p-nitrophenyl phosphate substrate was added to each well. The developed color was read at 405 nm. TNF assay. RAW 264.7 cells were maintained in DMEM medium and exposed to ET or to ET derivatives. The supernatant was added to L929 target cells in various dilutions. Cytotoxicity was
Potentiation of HIV Envelope Glycoproteins measured by staining the surviving and still adhering L929 cells with crystal violet according to the method of Ruff & Gifford (1981). Determination o f endotoxicity. Chick embryo lethality (Smith & Thomas, 1956) and Shwartzman skin reactivity in rabbits were used to estimate the in vivo toxicity of the isolates according to the methods described elsewhere (Nowotny, 1979). The chromogenic LAL kit of Whittaker M.A. Bioproducts (Walkersville, MD) was used to quantitate ET content according to the instructions of the manufacturer. The ET content was expressed as ET units. Chemical analysis. The percent total reducing carbohydrate, hexosamine, carboxylic acid and amino acid content of the samples was carried out as described before (Nowotny, 1979). Qualitative analysis of the carboxylic acids was carried out by GLC of their methyl esters (Nowotny, 1979). For g a s - l i q u i d chromatography of fatty acids, 1 mg samples were transesterified using BF3 in anhydrous methanol as described elsewhere (Nowotny, 1979). The hexane extract of the fatty acid methyl esters was used to quantitate the ester content by the hydroxylamine procedure. Five microliter atiquots were injected into a Hewlett Packard (Palo Alto, CA) 5830-A GLC for the identification of the methyl esters. Chromatographic conditions: 10070 SP-2330 on 100-120 mesh Supelcoport column (Supelco, Inc., Bellefonte, PA). The temperature was 100- 220°C, with an 8°C/min gradient. Dry-weight determination was done by pipetting 100~1 aliquots onto a pre-weighed 20 x 20 mm aluminum foil piece and drying under an infrared lamp oven and weighed using an ultra microbalance. If the sample contained salts or buffers, the extracts were dialyzed exhaustively prior to dry-weight content determination.
RESULTS
Effect o f acidic hydrolysis on the adjuvant activity (ADA)
The ADA of hydrolyzates of S. marcescens ATCC No. 13477 ET with 0.2 N acetic acid at 95°C is shown in Fig. 1. The immunogen in these experiments was freshly drawn sheep red blood cell (SRBC) suspension. The value 1.0 is the immune response elicited by SRBC alone. It is evident from this figure that short hydrolysis produces a potent Lipid A precipitate and a PS which is inert. Longer hydrolysis apparently released adjuvant active
575
SERRATIA MARCESCENS A'B'CC 13477
,.o
I I
I
I 3.0
I PS
.~_ --~ ==
2.0
1.0
• SRBC only
I
I
0'
I
I
I
I
30' 60' 90' 120'
240'
Time of hydrolysis (min)
Fig. 1. Immune response to SRBC in mice. Effect of continuous mild hydrolysis on the immunomodulating
activity of ET from S. marcescens ATCC 13477.
SERRATIA
MARCESCENS
ATTCC
13477
4.0
\LIPIDA 2.0-
1.0 .
.
0'
I
.
I
.
I
I
CBre3 only
I
30' 60' 90' 120' 240' Time of hydrolysis (min)
Fig. 2. Immune response to recombinant HIV glycoprotein CBre3 in mice. Effect of continuous mild hydrolysis on the immunomodulating activity of ET from S. marcescens ATCC 13477. Values less than 1.0 stimulation index mean immunosuppression. components into the supernatant, and this release reaches its maximum around 120 min. Continued hydrolysis destroys the adjuvant activity of the PS components. The same S. marcescens ATCC 13477 ET and its split products were also tested for adjuvant activity
576
E. KOVATS el al. Serratla
marcescens
ATCC
13477
SAMONELLA
4.0-
3.0-
'° 1
/
MINNESOTA
1114
-. j,
3.0-
._~ E
N 20
2.0-
2
1.0-
GLA 40 only
1.0
CBre3 only
I
I
0' Time of hydrolysis (rain)
Fig. 3. Immune response to random synthetic polypeptide GLA-40 in mice. Effect of continuous mild hydrolysis on the immunomodulating activity of ET from S. marceseens ATCC 13477. in c o m b i n a t i o n with CBre3 a n d G L A - 4 0 polypeptide i m m u n o g e n s . T h e results were similar to those observed using S R B C as the i m m u n o g e n . S h o r t hydrolysis p r o d u c e d a p o t e n t Lipid A a d j u v a n t a n d an inert, or in some cases, an a p p a r e n t l y i m m u n o s u p p r e s s i v e PS. If the hydrolysis was c o n t i n u e d for 60 a n d 120 m i n the PS b e c a m e active but f u r t h e r hydrolysis destroyed the a d j u v a n t activity o f the PS. In some cases it restored the i m m u n e e n h a n c i n g p o w e r o f the Lipid A. Figures 2 a n d 3 show these findings. If S. m i n n e s o t a S 1114 E T was subjected to identical hydrolysis, a n d samples were t a k e n for A D A d e t e r m i n a t i o n d u r i n g the time course o f the e x p e r i m e n t , the p a t t e r n o f the changes o f A D A was s o m e w h a t different f r o m those s h o w n in Figs 1, 2 a n d 3 for S. m a r c e s c e n s A T C C 13477. T h e Lipid A p r e p a r a t i o n s were m o r e p o t e n t t h a n the PS samples d u r i n g the first 2 h o f hydrolysis, but the PS gained A D A if the hydrolysis was c o n t i n u e d for 4 h. T h e Lipid A also showed increasing A D A d u r i n g the same time. Figure 4 presents the SI values for CBre3 as a n i m m u n o g e n .
I
I
~
I
30' 60' 90' 120' 240' Time of hydrolysis (rain)
Fig. 4. Immune response to recombinant HIV glycoprotein CBre3 in mice. Effect of continuous mild hydrolysis on the immunomodulating activity of ET from S. minnesota 1114.
Table 1. Polyclonal B-cell activation by S. marcescens ATCC 13477 preparations Preparations*
Increase of anti-CBre3 immunoglobulin levels ~
ET
2.7
30 60 120 240
PS min min min min
95°C 95°C 95°C 95°C
1.3 1.2 1.2 0.9
30 60 120 240
Lipid A min 95°C min 95°C min 95°C min 95°C
2.4 2.1 1.6 1.2
*10 Mg were injected i.p. into CD-I mice. *The anti-CBre3 level of saline and of preparation injected mouse sera were determined by ELISA. The numbers ion this column indicate the x fold increase in the anti-CBre3 globulin levels.
Polyclonal B-cell activations
M o s t ET p r e p a r a t i o n s tested so far showed some activity in releasing CBre3-reactive i m m u n o g l o b u l i n s in vivo. T h e Lipid A samples were also active; in a
few cases as active as the ET samples but the PS showed insignificant P B A i n d u c t i o n (Tables 1 a n d 2).
Potentiation of HIV Envelope Glycoproteins Table 2. Polyclonal B-cell activation by S. minnesota 1114 and preparations Preparations*
Increase of anti-CBre3 immunoglobulin levels~
ET Sl114 R595
2.2 2.7
PS* min 95°C min 95°C min 95°C min 95°C
1.0 0.6 1.6 0.2
Lipid A* 30 min 95°C 60 min 95°C 120 min 95°C 240 min 95°C
2.8 2.3 0.4 NT
30 60 120 240
*10 tag were injected i.p. into CD-1 mice. +The anti-CBre3 level of saline and of preparation injected mouse sera were determined by ELISA. The numbers in this column indicate the x fold increase in the anti-CBre3 immunoglobulin levels. *Preparations obtained from S. minnesota S1114.
Table 3. Release of total serum IgG by S. minnesota 1114 preparations Preparations*
Increase of total serum IgG levelst
Saline
1.0
ET S1114 R595
10.5 2.9
PS* 30 min 60 min 120 min 240 min
8.6 2.7 7.1 0.9
Lipid A* 30 min 60 min 120 min 240 min
10.1 8.0 1.8 NT
*10 ~g were injected i.p. into CD-1 mice. tThe IgG level of saline and of preparations injected mouse sera were determined by ELISA. The numbers in this column indicate the x fold increase in the immunoglobulin levels. *These preparations were obtained from S. minnesota S1114.
577
Release o f serum IgG T h e t o t a l IgG level in the sera of the injected mice was o f t e n m a n y fold increased as c o m p a r e d with the saline injected CD-1 mice. The m o s t active were S. minnesota S 1114 ET, the 30 a n d 120 m i n PS, the 30 a n d 60 m i n Lipid A f r o m the same s t r a i n ' s E T (Table 3). Some o f these p o t e n t a d j u v a n t , P B A active or IgG level e n h a n c i n g p r e p a r a t i o n s were n o longer endotoxic (Table 4).
T N F and L A L activities o f ET, Lipid A and PS T a b l e 4 presents the results o f T N F a n d L A L activity changes d u r i n g acidic hydrolysis. G o o d c o r r e l a t i o n has been f o u n d between the L A L positivity a n d T N F inducing capacity of the fractions. Neither the T N F n o r L A L activity correlated with the immune responsiveness e n h a n c i n g potency o f the PS samples.
Endotoxicity o f the hydrolyzed preparations Table 4 also shows the changes in chick e m b r y o lethality a n d in the local S h w a r t z m a n reaction d u r i n g hydrolysis with 0.2 N acetic acid. T h e endotoxicity o f the samples rapidly diminishes d u r i n g such treatment.
Chemical analysis T a b l e 5 shows the c o m p o s i t i o n o f S. marcescens A T C C 13477 ET, a n d PS extracts. The c o m p o n e n t s o f the E T are similar to those reported for E T p r e p a r a t i o n s isolated by the T C A p r o c e d u r e o f Boivin a n d associates. The analysis o f the PS o b t a i n e d by 0.2 N acetic acid cleavage at 95°C in 120 m i n shows two m a j o r changes as c o m p a r e d with the starting material (ET) : the e l i m i n a t i o n o f fatty acids a n d p h o s p h o r u s . A p p r o x i m a t e l y 80°7o o f the fatty acid loss occurs in the first 30 min o f hydrolysis (data not shown).
G L C o f fatty acids The fatty acid c o n t e n t of S. marcescens E T samples varied between 16 a n d 17°70. PS p r e p a r a t i o n s o b t a i n e d after 120 or 240 rain hydrolyzes at 95°C in 0.2 N acetic acid c o n t a i n only trace a m o u n t s o f fatty acid ( < 0 . 5 % ) . Figure 5a shows the fatty acid c o n t e n t o f S. mareescens E T a n d Fig. 5b shows t h a t o f a 120 m i n PS sample, carried o u t u n d e r identical conditions.
578
E. KOVATS et al.
Table 4. Changes in the TNF generating and endotoxic activity of the preparations during acidic hydrolysis. Preparations
TNF*
LAL +
SHWM*
LD50 ~
100 100 --
1.8 x 106 3.4 X 1 0 7 6.9 × 107
4.5 3.8 4.1
0.015 0.022 0.080
ET S. marcescens S. minnesota Sl114 S. minnesota R595
PS S. marcescens
30 60 120 240
min min min min
Neg. ----
2.8 × 10~ 0.3 × 10' Neg. Neg.
1.4 0.3 Neg. Neg.
0.92 5.00 10.00 10.00
30 60 120 240
min min min min
-----
3.5 × 105 3.3 × l0 s Neg. Neg.
2.5 1.3 0.3 Neg.
1.23 5.00 10.00 10.00
30 60 120 240
min min min min
52 32 24 12
9.6 4.7 2.8 0.2
× × × ×
105 105 105 103
3.2 2.2 2.9 2.5
0.09 0.03 0.10 0.25
30 60 120 240
min min min min
-----
5.5 4.2 3.3 7.4
× × >( ×
107 107
2.9 2.7 2.0 0.8
0.02 0.02 0.15 0.09
S. minnesota 1114
Lipid A S. marcescens
S. minnesota SI 114
10 7
105
*TNF units/ml are expressed according to the procedure of Ruff & Gifford (1981). +Limulus lysate chromogenic assay results expressed as ET units per/ag preparations.
*Local Shwartzman reaction. Numbers show size of hemorrhagic area in square cm. ~Chick embryo lethality LDs0, in ~g. (--) Not tested. (Neg.) No reaction. DISCUSSION To study the molecular aspects o f i m m u n e p o t e n t i a t i o n by ET and its analogs our a p p r o a c h is the identification o f the minimal structural requirements in the macromolecule o f ET or its c o n t a m i n a n t s , which are able to induce i m m u n e enhancement. The use o f fragments for the identification o f A D A active structural elements o f ET might need explanation, particularly these days when synthetic c o m p o u n d s with c o m p a r a b l e endotoxicity are available. First o f all, these synthetic c o m p o u n d s could be synthesized only after the structure o f the lipid moiety had been elucidated by a n u m b e r o f earlier investigators. The m o n o p h o s p h o r y l lipid A ( M P L ) p r e p a r a t i o n s are p o t e n t adjuvants, but they
are complex mixtures o f n u m e r o u s split p r o d u c t s and it has not been firmly established which one o f these or which c o m b i n a t i o n o f these is required for A D A . As will be discussed below, our adjuvant active PS preparations do not show any resemblance to Lipid A c o m p o n e n t s or M P L . They represent a new structural entity o f ET and A D A . The aim o f our studies, therefore, is first the exact identification o f the structure o f the i m m u n o p o t e n t i a t i n g PS c o m p o n e n t , and if we succeed with these efforts, the next step will be a t t e m p t s to synthesize such components. As far as the i m m u n o g e n s are concerned, we selected three: the convenient SRBC, a synthetic r a n d o m polypeptide GLA-40 and a r e c o m b i n a n t HIV-1 envelope protein, CBre3. We carried out preliminary experiments to establish the dose o f the
Potentiation of HIV Envelope Glycoproteins Table 5. Chemical analysis of ET and PS preparations Constiuents %
ET*
PSi
Amino acids* Amino sugars ~ Reducing carbohydrates H KdoI Fatty acids** Phosphorus t* Nucleic acids**
8.14 14.32
3.21 18.10
31.8 1.46 16.06 1.02 2.30
46.96 1.57 0.21 0.05 3.75
*TCA extracted and purified ET from S. marcescens ATCC 13477. *PS from above ET obtained by hydrolysis with 0.2 N acidic acid at 95°C for 120 min. *Expressed a % glycine, bassed on primary amine determination and corrected for hexosamines. ~Expressed as % D-glucosamine. "Expressed as % D-glucose. I°70 2-keto-3-deoxy octulosonic acid. **Expressed as % palmitic acid. t'Total % phosphorus content. t'Total % nucleic acid content.
11.83 14.15 .9:, I /16.12
B
li 21.10
i
l
21.10
1.1_ Fig. 5. GLC of fatty acids. (A) Fatty acid methylesters in TCA extracted and purified ET from S. marcescens ATCC 13477. (B) Same analysis carried out on a 120 rain 95°C hydrolyzed PS preparation. immunogens so that a low but definite immune response would be elicited without adjuvants. In the experiment reported here, the adjuvants were given at the same time as the immunogen but by separate injections, and both were injected i.p. These
579
experimental conditions leave a number of other variables to be tested, such as variations of the ratio between the dose of immunogen and immune stimulator and the time interval between the injection of the adjuvant and the immunogen. Since the adjuvant potentials of the PS were reported earlier by us, it may be necessary to underscore the information in this publication which is unquestionably new. The first is that potent immune enhancers are among the partial hydrolytic products of ET, and that these do not elicit any of the effects of ET which are considered to be undesirable. With the exception of 30 min PS preparations, these adjuvant active isolates were negative even in the lymphocyte proliferation (mitogenicity) test (Frank et al., 1977; Friedman et al., 1988). Our earlier claim regarding the immunopotentiating effect of PS preparations has been questioned by the more traditional investigators of structure and function relationship in endotoxins. We believe that these new data give further substantiation to our claim by showing that several PS preparations have a definite immune adjuvant effect, tested on three different immunogens. The second point for discussion deals with the possible origin of the molecular fragments with adjuvant activity. The rather obvious question is whether they are split products of the lipid moiety or of the polysaccharide regions or non-endotoxic contaminants of the TCA extracted ET. Chemical analysis has been carried out so far only on a few samples, which showed the highest ADA. These limited data show rather convincingly that the PS do not contain residual ET. The fact that one of the most active PS samples does not contain more than trace amounts of fatty acid and phosphorus seem to indicate that these compounds are not contaminated with Lipid A either. It remains to be investigated whether these preparations are split products of the core region and whether they may contain deacylated fragments of the glucosamine backbone of the lipid moiety. Our chemical analysis also rules out the possibility that the PS preparations would contain " m o n o phosphoryl E T " (MPL) molecules (Qureshi, Takayama & Ribi, 1985). MPL is obtained by very mild acidic hydrolysis of mutant ET (DPL) and it is claimed that this treatment cleaved only the glycosidically bound phosphate group leaving the fatty acids intact rendering the ET non-toxic but still an adjuvant (Johnson & Tomai, 1990). We detected almost no phosphorus or fatty acid in PS preparations.
580
E. KOVATSet al.
We think that another noteworthy observation was the generation of CBre3-reactive immunoglobulins into the sera of mice which received adjuvants but no antigen. The immunochemical characterization of these immunoglobulins and particularly their specificity for the H I V envelope remain to be elucidated. Differences were found between S. marcescens A T C C 13477 and S. m i n n e s o t a S 1114 ET derived PS and Lipid A preparations tested as adjuvants. The unique patterns o f Figs 1, 2 and 3 show interesting changes in the adjuvant activity of the PS and Lipid A samples during acidic hydrolysis. Similar, but not identical, changes could be observed during the hydrolysis of S. m i n n e s o t a ET. This nonidentity might be attributed to either structural differences between the two ET preparations, or to the presence of components in the S. marcescens ET which were not present in the S. m i n n e s o t a sample. Since we found significant differences among the A D A of a few Gram-negative strains ET or PS, one should expect that similar or greater differences will be found if one chooses other Gram-negative strains for similar studies. Whether our findings are applicable to all G r a m negative bacteria and their products remains to be investigated. These experiments we report here were carried out by giving one i m m u n o g e n and one adjuvant injection
and the mice were sacrificed 7 days later. A number of further variations are carried out at the present time in our laboratory to optimize the A D A of our samples. These include repeated injections of CBre3 a n d / o r PS, variations of the time intervals between these two and measurements of the duration of the elevated anti-CBre3 levels in the sera of the mice. The possible existence of immunosuppressive structural elements in the ET samples we studied needs rigorous substantiation. It was shown unequivocally that the PS obtained after short (30 min) hydrolysis is either inert as an adjuvant or in some systems immunosuppressive as shown in Figs 1 and 2. Under the same conditions the 30 min Lipid A precipitate was active but became inert or slightly suppressive after 60 min. We decline to speculate at this time about the possible explanations of the drastic changes in A D A of the PS or Lipid A samples occurring during continued acidic hydrolysis but we pursue this observation by attempting to isolate immune suppressive split products during such hydrolytic processes. Acknowledgements - - This work has been supported by
NIH grants AI25348, CA24628 and by a donation received from Cell Technology, Inc., Boulder, CO. The authors gratefully acknowledge the provision of CBre3 recombinant HIV glycoprotein by Cambridge Bioscience, Worcester, MA.
REFERENCES
APTE, R. M., HERTOGS, C. F. & PLUZNIK, D. H. (1977). Regulation of lipopolysaccharide-induced granulopoiesis and macrophage formation by spleen cells. I. Relationship between colony-stimulating factor release and lymphocyte activation in vitro. J. Immun., 118, 1435 - 1440. BEHEING, U. H. & NOWOTNY, A. (1977a). Immune adjuvancy of lipopolysaccharide and a nontoxic hydrolytic product demonstrating oscillating effect with time. J. Immun., 118, 1905 - 1907. BEHLING, U. H. & NOWOTNY,A. (1977b). Long-term adjuvant effect of bacterial endotoxin in prevention and restoration of radiation-caused immuno-suppression. Proc. Soc. exp. Biol. Med., 157, 348- 353. BEHLING, U. H. & NOWOTNY,A. (1980). Cyclic changes of positive and negative effects of single endotoxin injections. In Bacterial Endotoxin and Host Responses (ed. Agarwal, M. K.), pp. 11-26. Elsevier/North Holland Biomedical Press, Amsterdam. BEHLING, U. H. & NOWOTNY,A. (1982). Bacterial endotoxins as modulators of specific and non-specific immunity. In Regulatory Implication o f Oscillatory Dynamics in the Immune Response (eds Hiernaux, J. and DeLisi, C.), Vol. 2, pp. 127- 140. CRC Press, Boca Raton, FL. BOIVIN, A., MESROBEANU,J. & MESROBEANU,L. (1933a). Technique pour la preparation des polysaccharides microbiens specifiques. Compt. Rend. Soc. Biol., 113, 490-492. BOIV1N, A., MESROBEANU,J. & MESROBEANU,L. (1933b). Extraction d'un complexe toxique et antigenique a partir du bacille d'Aertrycke. Comp. Rend. Soc. Biol., 114, 307-310. CHANG, H., THOMPSON, J. J. t~£ NOWOTNY, A. (1974). Release of colony stimulating factor (CSF) by non-endotoxic breakdown products of bacterial lipopolysaccharides. Immun. Commun., 3, 401-409. FRANK, S., SPECTER, S., NOWOTNY, A. • FRIEDMAN, H. (1977). Immunocyte stimulation in vitro by nontoxic bacterial lipopolysaccharide derivatives. J. Immun., 119, 855-869.
Potentiation of HIV Envelope Glycoproteins
581
FRIEDMAN, H., BUTLER, R. C., SPECTER, S. & NOWOTNY, A. (1988). Nontoxic endotoxin polysaccharide induces soluble mediators which potentiate antibody production by murine retrovirus suppressed splenocytes. Int. J. Immunopharmac., 10, 283 - 292. HAEEFNER-CAVAILLON, N., CAVAILLON, J. M. & SZABO, L. (1982). Macrophage-dependent polyclonal activation of splenocytes by Bordetella pertussis endotoxin and its isolated polysaccharide and lipid A regions. Cell. Immun., 74, 1-13. HAEFFNER-CAVAILLON,N., CAVAILLON,J. M. & SZABO, L. (1984). Interleukin 1 secretion by human monocytes stimulated by the isolated polysaccharide region of Bordetella pertussis endotoxin. Molec. Immun., 21, 389- 395. JOHNSON, A. G. & TOMAI, M. A. (1990). A study of the cellular and molecular mediators of the adjuvant action of a nontoxic monophosphoryl Lipid A. Adv. exp. Med. Biol., 256, 567-579. LUDERITZ, O. (1970). Recent results on the biochemistry of the cell wall lipopolysaccharides of Salmonella bacteria. Angew. Chem., 9, 649-663. NOWOTNY, A. (1979). Basic Exercises in Immunochemistry, 2nd Edn. Springer, New York. NOWOTNY, A., BEHLING, U. H. & CHANG,H. L. (1975). Relation of structure to function in bacterial endotoxins. VIII. Biological activities in a polysaccharide-rich fraction. J. Immun., 115, 199-203. QURESHI, N., TAKAYAMA,K. & RIBI, E. (1982). Purification and structural determination of nontoxic Lipid A obtained from the lipopolysaccharide of Salmonella typhymurium. J. biol. Chem., 257, 11808- 11815. RAICHVARG, D., BROSSARD, C. & AGNERAY, J. (1980). Preparation of a nontoxic and immunogenic polysaccharide fraction from a Haemophilus influenzae phenol-water extract. Infect. Immun., 29, 171 -174. RUFF, M. R. & GIFrORD, G. E. (1981). Rabbit tumor necrosis factor: mechanism of action. Infect. Immun., 3 1 , 3 8 0 - 385. SMITH, R. T. & THOMAS, L. (1956). The lethal effect of endotoxin on the chick embryo. J. exp. Med., 104, 217-231. THORN, R. M., BELTZ, G. A., HUNG, C., FALLIS, B. F., WINKLE, S., CHENG, K. & MARCIANI,D. J. (1987). Enzyme immunoassay using a novel recombinant polypeptide to detect human immunodeficiency virus env antibody. J. cfin. Microbiol., 25, 1207- 1212. URBASCHEK,R. (1980). Effects of bacterial products of granulopoiesis. Adv. exp. Med. Biol., 121b, 5 1 - 64. WESTPHAL, O. & LUDERITZ,O. (1954). Chemische Erforschung yon Lipopolysacchariden Gram-negativer Bakterien. Angew. Chem., 66, 407-417. WESTPHAL,O. & EUDERITZ,O. (1960). 3.6-didesoxy-hexosen-Chemie und Biologie. Angew. Chem., 72, 881- 891.
