Saturday PHAGOCYTOSIS IN PATIENTS AND CARRIERS OF CHRONIC GRANULOMATOUS DISEASE
W. DOUGLAS BIGGAR
Department of Pediatrics and Immunology, Research Institute of the Hospital for Sick Children,
Toronto, Ontario, Canada The
of
capacity leucocytes Summary from phagocytic four patients and two carriers of chronic granulomatous disease (C.G.D.) was compared that of normal leucocytes. Leucocytes from C.G.D. patients phagocytised more Staphylococcus aureus to
than did normal leucocytes after 5, 10, and 20 minutes of incubation. Phagocytosis by leucocytes from two carriers of C.G.D. was intermediate between that in the patients and in controls. In contrast, phagocytosis of Streptococcus fœcalis, an organism readily killed by C.G.D. leucocytes, was similar for control and C.G.D. leucocytes. Enhanced phagocytosis may represent an attempt by leucocytes to compensate for a bactericidal abnormality. In addition, these observations may partly explain why carriers of C.G.D. are not more susceptible to infection despite abnormal leucocyte function. Detection of carriers using the described assay of phagocytosis appears relatively simple compared to previously described methods.
Introduction VIGOROUS phagocytosis and killing of microorganisms by leucocytes are essential for host resistance against microbial disease. Considerable progress in dissecting the complex interactions between leucocytes and microorganisms has been made by studying leucocytes from patients with an increased susceptibility to bacterial infection. Indeed, many of the metabolic requirements for bacterial killing by leucocytes have been defined by studying leucocytes from patients with chronic granulomatous disease (C.G.D.). These studies have shown that for leucocytes increased cellular respiration, increased hexosemonophosphate pathway activity, and increased hydrogen-peroxide (H202) formation are normally associated with phagocytosis and contribute, in a major way, to the bactericidal capacity of the leucocyte.1,2 While the precise metabolic defect in leucocytes from patients with C.G.D. remains unknown, the impaired production of H202 appears to be of central importance for the observed abnormality in bacterial killing by such leucocytes.3 Other patients with recurrent infection and abnormal leucocyte 7914
3
May 1975
metabolism have been described who, in contrast to patients with C.G.D., appear less susceptible to bacterial infection.4 Despite their abnormality of leucocyte metabolism, these patients are usually free of infection. This suggests that the leucocyte may have alternative bactericidal mechanisms to compensate for deficiencies. Indeed, Klebanoff and Pincushave demonstrated in vitro that leucocytes deficient in the enzyme myeloperoxidase appear to compensate for this deficiency by utilising alternative, non-peroxidasemediated, bactericidal mechanisms. Thus, these invitro analyses of leucocyte function support clinical observations which suggest that leucocytes may have an overkill capacity and alternative bactericidal mechanisms to compensate when one bactericidal mechanism is
compromised. Phagocytosis is another facet of leucocyte function which might be influenced by changes in the bactericidal capacity of leucocytes. Until recently, it has been difficult to study precisely the early kinetics of phagocytosis. Previously, many techniques have been used to try and dissect the phagocytic and bactericidal events in leucocytes. To achieve this, one must be able to identify and to quantitate extracellular bacteria, adherent to phagocytic cells but not yet ingested, from extracellular bacteria. Methods to separate intracellular bacteria from extracellular bacteria have included differential centrifugation and the addition of antibiotics to the incubation mixture to kill the extracellular bacteria. Unfortunately, these methods have not permitted one to examine, in a critical manner, phagocytosis and killing by leucocytes as independent cellular events. Previous studies of leucocytes from patients with C.G.D. using these techniques have suggested that the phagocytic capacity of such leucocytes was normal or slightly impaired.3 The need to quantitate these events more precisely and, in particular, to examine the kinetics early in phagocytosis has been stressed.3,6.7
Lysostaphin,s a muralytic enzyme, rapidly eliminates extracellular Staphylococcus aureus and does not penetrate leucocytes.9 Recently, this enzyme has been used to examine phagocytosis by normal leucocytes in a more precise manner than earlier methods had permitted 1° While the intracellular events associated with the bactericidal defect in leucocytes from patients with C.G.D. have been studied extensively, the phagocytic capacity of leucocytes from these patients has not been well documented. Such information might improve both our understanding of this disease and our appreciation of the complex interaction between bacteria and leucocytes. s
992
compared the phagocytic capacity of normal leucocytes with that of leucocytes from patients with C.G.D. and leucocytes from known carriers of C.G.D. using an assay of phagocytosis which appears to be more precise than methods described previously. I have
Materials and Methods Patients Four male children with C.G.D. were studied when they clinically free of infection. Their ages ranged from 7 to 12 years and each patient had a history of recurrent and severe bacterial infection. The diagnosis of C.G.D. was established in each instance by demonstrating the typical bactericidal abnormality of their leucocytes 11 In addition, leucocytes from each patient failed to show the increased oxygen consumption, increased hexose-monophoswere
phate pathway activity, and increased tetrazolium-dye reduction normally associated with phagocytosis.l,2 When possible, family members were studied for the presence of carriers. A carrier was defined as a female whose leucocytes were intermediate in their bactericidal capacity and in whom two populations of neutrophils could be demonstrated using a histochemical test of nitroblue-tetrazolium (N.B.T.) dye reduction.12,13
Preparation of Leucocytes Leucocyte-rich plasma was obtained by sedimentation of heparinised venous blood in a solution of 6% dextran in saline (4 ml. dextran and 100 units of heparin per 20 ml. of blood). The leucocytes were washed twice in Hanks’ balanced salt solution (H.B.s.s.), and resuspended in H.B.s.s. to contain 107 leucocytes per ml. Leucocytes prepared in this way regularly contained 60-80% neutrophils as determined microscopically.
Leucocytes at
were
preincubated with the drug for 20 minutes
37°C.
In some experiments, the degree of phagocytosis was also assessed morphologically after 10 and 20 minutes of incubation. Smears of the leucocyte/bacteria suspension were prepared by cytocentrifugation and stained with methyl-green pyronine. At least 200 neutrophils were examined microscopically for the presence and absence of cell-associated bacteria. The number of bacteria per cell was calculated for those leucocytes which appeared to have ingested bacteria.
Assay of Phagocytosis Using Differential Centrifugation In these experiments, leucocyte/bacteria suspensions After identical periods of were prepared as described. incubation, 0’2 ml. of the mixture was removed and washed twice in H.B.S.S. to remove the extracellular bacteria. The cell pellet was lysed in distilled water and the viable cell-associated bacteria were quantitated as described. No antibiotics were added to the assay.
Results
Phagocytic Capacity ofLeucocytes The phagocytic capacity of leucocytes from four patients with C.G.D., from two known female carriers of C.G.D., and from controls was assessed using lysostaphin (fig. 1). In the experiments shown, 5-0 X 101 neutrophils were tested with 106 bacteria and aliquots were examined after 5, 10, and 20 minutes of incubation. As shown in fig. 1, the capacity of the patients’ leucocytes to phagocytise Staph. aureus exceeded the phagocytic capacity of control leucocytes. In experiments not shown here, similar results were obtained when the bacteria/leucocyte ratio was varied by
Preparation of the Test Organisms Staph. aureus 502A and Streptococcus faecalis were used as test organisms after an 18-hour culture. Just prior to use, bacteria were washed three times in physiological saline solution and resuspended in physiological saline to contain approximately 107 organisms per ml. The precise number of bacteria was assessed for each experiment using standard dilution and pour-plate techniques.
Assay of Phagocytosis Using Lysostaphin The method used
examine the kinetics of early phagocytosis by leucocytes was modified from the method described by Tan et al.1o Briefly, 0-5 ml. of the leucocyte suspension, 0’1 ml. pooled normal human serum (P.N.H.s.), 0’1 ml. bacterial suspension, and 0’3 ml. H.B.s.s. were added to a sterile 12 X 75 mm. disposable Falcon tube and the mixture was incubated at 37°C on an aliquot mixer (Lab-Tek, Inc., Westmont, Illinois). Control tubes, one without leucocytes (bacteria control) and the other without serum, were examined in parallel. After 5, 10,’ and 20 minutes of incubation, a 0-2 ml. aliquot was removed from each tube. To each aliquot was added 3’5 units of lysostaphin (Mead Johnson Research Center, Evansville, Ind.), and the mixture was incubated at 37°C for 20 minutes. To inactivate the lysostaphin, 0-015 ml. of 2-5% trypsin isotonic saline solution was added and the mixture was incubated for 10 minutes. The incubation mixture was then diluted with 1’75 ml. sterile distilled water to lyse the leucocytes, and the number of viable intracellular bacteria was quantitated using standard dilution and pour-plate techniques. In several experiments, different bacteria to neutrophil ratios were examined (1/5 to 100/1). Since several drug-induced abnormalities of leucocyte function have features similar to the leucocyte abnormalities described in C.G.D.,14,15 the kinetics of phagocytosis was assessed in the presence of 2’5xlO’M phenylbutazone. to
phagocytosis of Staph. aureus by leucocytes patients with C.G.D. (.———A), carriers of C.G.D. (D———D). and controls (0-0), using lysostaphin.
Fig. 1-Early from
The IOgl. number viable bacteria per ml. is plotted versus time in minutes. Data represent the means.E. of eleveo experiments for patients and controls and the range of two experiments for two carriers. Rate of killing Staph. aureus by lysostaphin ( - - - - W) and bacteria control (N - - - -N).
993
and the variability of the assay. While this technique was less sensitive than the lysostaphin method, which
neutralised extracellular Staph. aureus, differential centrifugation could detect differences, although much smaller, in early phagocytosis of Staph. aureus between leucocytes from C.G.D. patients and controls. In separate experiments in this laboratory (data not shown) and by Holmes and GOOd,14 leucocytes from C.G.D. patients killed Strep. f (zcalis
rapidly
normally.
,
Influence of Phenylbutazone on Phagocytosis The phagocytic capacity of normal leucocytes for Stap7i. aureus in the presence of 2.5 X 10--’M phenyl-
Fig. 2-Representative experiment comparing early phagocytosis of Strep. feecalis by leucocytes from patients with C.G.D. (A.——A ), carriers of C.G.D. ( D—— D): and controls ( 0—— 0). Using differential centrifugation, the loglo number viable bacteria is plotted versus time in minutes.
butazone was increased 5-10 fold when tested in four experiments. By contrast, phenylbutazone had no effect on the phagocytic capacity of C.G.D. leucocytes. In parallel experiments (data not shown) and in studies by Holmes et aL14 similar concentrations of this drug induced a bactericidal abnormality in normal
leucocytes similar
to
that in C.G.D.
leucocytes.
Discussion
10-fold increments from 1 bacterium per leucocyte An increased phagoto 100 bacteria per leucocyte. for aureus Staph. cytic capacity by C.G.D. leucocytes was also suggested in experiments using differential centrifugation techniques, although the differences were much smaller. Further, when smears of leucocytes were assessed morphologically for phagocytosis after 10 and 20 minutes’ incubation with Staph. aureus, the number of bacteria associated with C.G.D. leucocytes appeared to be 2 to 5 fold greater than the number of bacteria associated with control leucocytes. It is also apparent in fig. 1 that phagocytosis by the patients’ leucocytes was maximum after only 5 minutes of incubation and did not appear to increase significantly with longer incubation. By contrast, with longer periods of incubation normal leucocytes continued to phagocytise bacteria, and at the times studied their phagocytic capacity did not equal the phagocytic capacity of C.G.D. leucocytes. In two experiments, leucocytes from two carriers of C.G.D. appeared to be intermediate in their capacity to phagocytise Staph. aureus. Patients’ sera did not influence the phagocytic capacity of normal leucocytes. Under the experimental conditions employed, no significant bacterial growth occurred for up to 3 hours of observation (fig. 1). A
representative experiment comparing the phagocytic capacity of leucocytes from the patients and controls for Strep. fcccalis is summarised in fig. 2. In these experiments, the phagocytic capacity of
leucocytes was assessed after similar times of incubation using differential centrifugation to separate the cell-associated bacteria from non-cell-associated, or supernatant, bacteria. As shown in fig. 2, phagocytosis of Strep. facalis by C.G.D. leucocytes was similar to that by control leucocytes. In four experiments, the rate of phagocytosis of Strep. facalis by leucocytes was less than with control leucocytes. This difference, although seemingly reproducible, may not be significant when one considers the sensitivity C.G.D.
Impaired antimicrobial activity of leucocytes can be due to abnormalities in the ability of leucocytes to ingest bacteria or to an intracellular microbicidal defect.16-18 Furthermore, microorganisms themselves appear to influence, in a major way, the extracellular and the intracellular phases of microbial inactivation.19 diminished bactericidal and metabolic releucocytes from patients with C.G.D. have well documented.’,’By contrast to the sophisticated metabolic studies, the methods of examining the kinetics of phagocytosis by leucocytes have been much less precise. Thus, the phagocytic capacity of leucocytes from C.G.D. patients, although generally reported to be normal or slightly diminished, has never been examined with a sensitive The
sponses of
tbeen
technique.3 The observations reported here demonstrate that the phagocytic capacity of leucocytes from patients with C.G.D. varies considerably with the test organisms and the assays used. Phagocytosis by leucocytes from patients with C.G.D. of Staph. aureus, an organism not readily killed by these cells, appears to be greater than phagocytosis by control leucocytes (fig. 1). Indeed, leucocytes from patients with C.G.D. phagocytised as many as 300-fold the number of Staph. aureus organisms which were phagocytised by control leucocytes. In these experiments, the time required by leucocytes from patients with C.G.D. to achieve maximum phagocytosis appeared to be very short. Leucocytes from patients with C.G.D., when tested at several different bacteria/leucocyte ratios, achieved maximum phagocytosis after only 5 to 10 minutes’ incubation. By contrast, the rate of early phagocytosis by normal leucocytes appeared to be slower. Quantitation of phagocytosis by normal leucocytes after periods of incubation greater than 20 minutes was difficult to assess since, after 20 minutes of incubation, a significant number of intracellular bacteria were being killed by the leucocytes. Many factors appear to influence the union of phagocytic cells with bacteria, both in vitro and
994 factors have been retard this process. can enhance or Indeed, a fragment of gamma-globulin has been reported to promote the phagocytosis of staphylococci by phagocytic cells z° A serum factor would not appear to account for the observations reported here, since P.N.H.S. and sera from patients with C.G.D. supported phagocytosis of control leucocytes and patient leucocytes equally. In addition, significant intracellular multiplication of bacteria in C.G.D. leucocytes would not account for the observations reported here since multiplication in the bacterial controls in each experiment was negligible (fig. 1). In order to examine these observations further, the phagocytic capacity of leucocytes from patients with C.G.D. was assessed using a test microorganism which their leucocytes could normally kill. These studies were important in order to determine if the phagocytic capacity of C.G.D. leucocytes for Staph. aureus was enhanced nonspecifically, or was influenced perhaps in some way by the ability of the leucocyte to kill the test microorganism. As shown in fig. 2, using a less sensitive assay of phagocytosis Strep. fœcalis, an organism which was killed vigorously by C.G.D. leucocytes,14. appeared to be ingested as readily by control leucocytes as by C.G.D. leucocytes. One possible explanation of these observations is that the phagocytic capacity of a leucocyte is influenced, in some way, by its ability to kill an ingested microorganism. Thus, the superior rate of early phagocytosis of Staph. aureus by C.G.D. leucocytes may represent an attempt by the patients’ leucocytes to compensate for their abnormal bactericidal capacity. Some drug-induced abnormalities of leucocyte function are similar to the bactericidal abnormality of C.G.D. leucocytes. 14 Although drugs are seldom selective in their influence on leucocyte metabolism, they have proven to be promising laboratory models of C.G.D. Phenylbutazone inhibits the bactericidal capacity of leucocytes for both catalasepositive and catalase-negative bacteria,14,15 but has no significant effect on phagocytosis. This laboratory model was then utilised to examine the possible effects of this drug on the early phagocytic capacity of normal leucocytes for Staph. aureus. In four experiments, phenylbutazone appeared to enhance the phagocytic capacity of normal leucocytes for Staph. aureus. This observation can be explained, in part, by the effect of phenylbutazone on bacterial killing by neutrophils. The bacteriostatic and bactericidal effects of neutrophils on ingested bacteria are expressed soon after the bacteria have been ingested and vary with the type of bacteria ingested. Although the bacteria may be unable to multiply, they may remain relatively intact and capable of continued cellular metabolism for longer periods.19 Active degradation of bacteria by leucocytes requires more time and again may vary considerably with the type of microorganism ingested. For staphylococci, with their inherent ability to resist both phagocytosis and killing by leucocytes, the leucocyte begins to digest the ingested microorganisms after approximately an hour.17 Differences between the bactericidal capacity of normal leucocytes and of C.G.D. leucocytes for Staph. aureus probably account for some of the observations reported in vivo.
In
vitro, several
described which
’""
serum
here. Normal leucocytes, by their early bacteriostatic effect on ingested bacteria, would have fewer viable intracellular bacteria, and would therefore appear to have phagocytised fewer bacteria than C.G.D. leucocytes which do not exert this bacteriostasis. This explanation, however, would not appear to account for the large difference in phagocytosis reported here since phagocytosis by leucocytes from patients with C.G.D., when assessed morphologically and by differential centrifugation, appeared to exceed that of control
leucocytes. Leucocytes
from female carriers of C.G.D. are intermediate between normal and leucocytes from patients with C.G.D. in their capacity to kill Staph. aureus in vitro. Approximately half of their peripheralblood leucocytes appear to have a metabolic defect similar to the C.G.D. patients’ 12,13 and half appear normal. However, this division between normal and abnormal cells may be quite variable as is illustrated by a female carrier described previously with bactericidal and metabolic abnormalities very similar to those in her brother with C.G.D.13 Despite these abnormalities of leucocyte function, which, in some carriers may be considerable,13 female carriers do not appear unduly susceptible to bacterial infection. While many more patients and carriers need to be studied, the two carriers studied in this report had an increased capacity to phagocytise Staph. aureus (fig. 1). This increased phagocytic capacity may function as an important compensatory mechanism by leucocytes in carriers of C.G.D. Such a mechanism in carriers might circumvent an increased susceptibility to bacterial infection and at least partly account for their clinical well-being. While more carriers of C.G.D. need to be studied, this assay of phagocytosis using lysostaphin appears to be useful for detecting C.G.D. carriers and to be relatively simple compared to previously described methods for the detection of C.G.D. carriers. These studies seem to demonstrate another aspect of leucocyte function and may provide some information on the recognition and controlling mechanisms of leucocytes. Clearly, greater understanding of the phagocyte cell-membrane and of the physical and biochemical properties of the surface of the microorganisms that influence particle ingestion is needed. I thank K. Y. Wong for her excellent technical assistance and Dr A. R. C. Cole and Dr P. D. McClure for permission to study their patients. This study was supported by the Medical Research Council of Canada (Grant MA5276). W. D. B. is a Medical Research Council scholar. Requests for reprints should be sent to W. D. B., Department of Immunology, Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada M5G 1X8. REFERENCES 1. Holmes, B., Page, A. R., Good, R. A. J. clin. Invest. 1967, 46, 1422. 2. Baehner, R. L., Nathan, D. G. Science, 1967, 155, 835. 3. Karnovsky, M. L. Fedn Proc. 1973, 32, 1527. 4. Holmes, B., Park, B. H., Biggar, W. D. in Molecular Pathology (edited by C. C. Thomas and S. B. Day) (in the press). 5. Klebanoff, S. J., Pincus, S. H. J. clin. Invest. 1971, 50, 2226. 6. Mandell, G. L., Hook, E. W. J. Bact. 1969, 100, 531. 7. Stossel, T. P., Mason, R. J., Hartwig, J., Vaughan, M. J. clin. Invest.
1972, 51, 615. 8. Schindler, C. A., Schuhardt, V. T. Biochem. 97, 242.
Biophys. Acta, 1965,
995
LYMPH-NODE BIOPSY DURING SIMPLE MASTECTOMY A. A. SHIVAS ELIZABETH L. M. CANT A. P. M. FORREST
Departments of Clinical Surgery and Pathology,
University of Edinburgh The distribution of pectoral (external mammary) nodes identified during the removed with the axillary tail of the and operation breast was studied in 45 patients treated by simple (total) mastectomy. Up to 13 nodes may lie within the axillary tail, and these are continuous with the pectoral nodes. Lymph-nodes were identified in 90% of patients treated by simple (total) mastectomy without dissection of the axilla. Sum ary
tail postoperatively; and "pectoral nodes" as those selected peroperatively by the surgeon (fig. 1). Methods
Simple mastectomy was performed in 45 patients with early breast cancer. Our technique for this procedure is described elsewhere.5 It includes careful definition and dissection of the axillary tail of the breast from between the pectoralis major muscle in front and the latissimus dorsi behind, and its removal with the breast up to the point where it blends with the axillary fat-i.e., at the level of the third rib. This is facilitated by removing the
Introduction WE have suggested that a rational policy for the local management of primary breast cancer is simple mastectomy with pectoral-node biopsy, followed by immediate postoperative radiotherapy only in those cases in which involvement of these nodes is proved histologically. Since these nodes can be identified without dissecting the axilla, the morbidity of either unnecessary axillary dissection or radical radiotherapy is avoided. We reported a controlled randomised study which was designed to assess this policy of treatment in Cardiff and St. Mary’s Hospital.’ This is now also the basic procedure used in the current Edinburgh breast cancer trials, which include the administration of additional systemic therapy to those with histologically involved pectoral lymph-nodes.3.4t We have been uncertain as to the true distribution of these pectoral nodes. Initially we believed that be near to the top of the best identified they could excised breast specimen, but later axillary tail of the we advised that the surgeon should identify a node or nodes of this group during his dissection of the axillary tail, particularly at the point of separation from the axillary fat. The study we now report was undertaken to determine the best method of accurate sampling of these nodes. It suggests that a careful search of the excised axillary tail and a careful search by the surgeon peroperatively are both essential steps for complete sampling. For clarity, we have defined " axillary-tail nodes " as those identified in the axillary
9. Schaffner, W., Melly, M. A., Hash, J. H., Koenig, M. G. J. biol. Med. 1967, 39, 215. 10. Tan, J. S., Watanakunakorn, C., Phair, J. P. J. Lab. clin. Med. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.
1971, 78, 316. Holmes, B., Quie, P. G., Windhorst, D. B., Good, R. A. Lancet, 1966, i, 1225. Windhorst, D. B., Page, A. R., Holmes, B., Quie, P. G., Good, R. A. J. clin. Invest. 1967, 47, 1026. Biggar, W. D., Buron, S., Holmes, B. J. Pediat. (in the press). Holmes, B., Good, R. A. J. reticul. Soc. 1972, 12, 216. Strauss, R. R., Paul, B. B., Sbarra, A. J. J. Bact. 1968, 96, 1982. Stossel, T. P. New Engl. J. Med. 1974, 290, 717. Quie, P. G. in Current Problems in Pediatrics, vol. XI, no. 11, p. 1. Chicago, 1972. Baehner, R. L. J. Pediat. 1974, 84, 317. Elsbach, P. New Engl. J. Med. 1973, 289, 846. Najjar, V. A., Constantopoulos, A. J. reticul. Soc. 1972, 12, 197.
Fig. 1-Position of axillary-tail and pectoral lymph-nodes. breast from the medial to lateral side. Separation of the tail from the axillary fat pad is facilitated by grasping it and feeling between finger and thumb the point at which the nodular breast fat merges into the much finer and smoother axillary fat lobules. After removal of the breast, the axillary tail was detached and the cut end marked with a suture. It was spread on a cork board and palpated on the flat. Palpable nodes were removed and their site marked on a sketch plan. The axillary tail was then sliced and any further nodes removed and also marked on the plan. Each node was fixed in 10% formol saline solution, embedded in paraffin, and sectioned. These nodes are termed " axillary-tail nodes ". Several sections were taken from each half of the node and stained with haematoxylin and eosin. When examination of these initial sections was negative, serial 1/10 sections were cut from the nodes, stained, and examined for foci of tumour. In all cases, sections were taken to confirm the presence of breast parenchyma in the axillarytail specimens. In 38 cases, at operation the surgeon also attempted to identify a node or nodes separate from the axillary tail These usually lay at the junction of the axillary tail and the axillary fat, and towards its medial aspect. A node was defined in 29 cases; 2 of these on section proved to be only fat. These nodes have been termed " pectoral
Results
Total Yield of Nodes from Axillary Tail Nodes were identified within the breast of the axillary tail in 32 (71 %) of the 45
parenchyma cases.
The
W. DOUGLAS BIGGAR
Department of Pediatrics and Immunology, Research Institute of the Hospital for Sick Children,
Toronto, Ontario, Canada The
of
capacity leucocytes Summary from phagocytic four patients and two carriers of chronic granulomatous disease (C.G.D.) was compared that of normal leucocytes. Leucocytes from C.G.D. patients phagocytised more Staphylococcus aureus to
than did normal leucocytes after 5, 10, and 20 minutes of incubation. Phagocytosis by leucocytes from two carriers of C.G.D. was intermediate between that in the patients and in controls. In contrast, phagocytosis of Streptococcus fœcalis, an organism readily killed by C.G.D. leucocytes, was similar for control and C.G.D. leucocytes. Enhanced phagocytosis may represent an attempt by leucocytes to compensate for a bactericidal abnormality. In addition, these observations may partly explain why carriers of C.G.D. are not more susceptible to infection despite abnormal leucocyte function. Detection of carriers using the described assay of phagocytosis appears relatively simple compared to previously described methods.
Introduction VIGOROUS phagocytosis and killing of microorganisms by leucocytes are essential for host resistance against microbial disease. Considerable progress in dissecting the complex interactions between leucocytes and microorganisms has been made by studying leucocytes from patients with an increased susceptibility to bacterial infection. Indeed, many of the metabolic requirements for bacterial killing by leucocytes have been defined by studying leucocytes from patients with chronic granulomatous disease (C.G.D.). These studies have shown that for leucocytes increased cellular respiration, increased hexosemonophosphate pathway activity, and increased hydrogen-peroxide (H202) formation are normally associated with phagocytosis and contribute, in a major way, to the bactericidal capacity of the leucocyte.1,2 While the precise metabolic defect in leucocytes from patients with C.G.D. remains unknown, the impaired production of H202 appears to be of central importance for the observed abnormality in bacterial killing by such leucocytes.3 Other patients with recurrent infection and abnormal leucocyte 7914
3
May 1975
metabolism have been described who, in contrast to patients with C.G.D., appear less susceptible to bacterial infection.4 Despite their abnormality of leucocyte metabolism, these patients are usually free of infection. This suggests that the leucocyte may have alternative bactericidal mechanisms to compensate for deficiencies. Indeed, Klebanoff and Pincushave demonstrated in vitro that leucocytes deficient in the enzyme myeloperoxidase appear to compensate for this deficiency by utilising alternative, non-peroxidasemediated, bactericidal mechanisms. Thus, these invitro analyses of leucocyte function support clinical observations which suggest that leucocytes may have an overkill capacity and alternative bactericidal mechanisms to compensate when one bactericidal mechanism is
compromised. Phagocytosis is another facet of leucocyte function which might be influenced by changes in the bactericidal capacity of leucocytes. Until recently, it has been difficult to study precisely the early kinetics of phagocytosis. Previously, many techniques have been used to try and dissect the phagocytic and bactericidal events in leucocytes. To achieve this, one must be able to identify and to quantitate extracellular bacteria, adherent to phagocytic cells but not yet ingested, from extracellular bacteria. Methods to separate intracellular bacteria from extracellular bacteria have included differential centrifugation and the addition of antibiotics to the incubation mixture to kill the extracellular bacteria. Unfortunately, these methods have not permitted one to examine, in a critical manner, phagocytosis and killing by leucocytes as independent cellular events. Previous studies of leucocytes from patients with C.G.D. using these techniques have suggested that the phagocytic capacity of such leucocytes was normal or slightly impaired.3 The need to quantitate these events more precisely and, in particular, to examine the kinetics early in phagocytosis has been stressed.3,6.7
Lysostaphin,s a muralytic enzyme, rapidly eliminates extracellular Staphylococcus aureus and does not penetrate leucocytes.9 Recently, this enzyme has been used to examine phagocytosis by normal leucocytes in a more precise manner than earlier methods had permitted 1° While the intracellular events associated with the bactericidal defect in leucocytes from patients with C.G.D. have been studied extensively, the phagocytic capacity of leucocytes from these patients has not been well documented. Such information might improve both our understanding of this disease and our appreciation of the complex interaction between bacteria and leucocytes. s
992
compared the phagocytic capacity of normal leucocytes with that of leucocytes from patients with C.G.D. and leucocytes from known carriers of C.G.D. using an assay of phagocytosis which appears to be more precise than methods described previously. I have
Materials and Methods Patients Four male children with C.G.D. were studied when they clinically free of infection. Their ages ranged from 7 to 12 years and each patient had a history of recurrent and severe bacterial infection. The diagnosis of C.G.D. was established in each instance by demonstrating the typical bactericidal abnormality of their leucocytes 11 In addition, leucocytes from each patient failed to show the increased oxygen consumption, increased hexose-monophoswere
phate pathway activity, and increased tetrazolium-dye reduction normally associated with phagocytosis.l,2 When possible, family members were studied for the presence of carriers. A carrier was defined as a female whose leucocytes were intermediate in their bactericidal capacity and in whom two populations of neutrophils could be demonstrated using a histochemical test of nitroblue-tetrazolium (N.B.T.) dye reduction.12,13
Preparation of Leucocytes Leucocyte-rich plasma was obtained by sedimentation of heparinised venous blood in a solution of 6% dextran in saline (4 ml. dextran and 100 units of heparin per 20 ml. of blood). The leucocytes were washed twice in Hanks’ balanced salt solution (H.B.s.s.), and resuspended in H.B.s.s. to contain 107 leucocytes per ml. Leucocytes prepared in this way regularly contained 60-80% neutrophils as determined microscopically.
Leucocytes at
were
preincubated with the drug for 20 minutes
37°C.
In some experiments, the degree of phagocytosis was also assessed morphologically after 10 and 20 minutes of incubation. Smears of the leucocyte/bacteria suspension were prepared by cytocentrifugation and stained with methyl-green pyronine. At least 200 neutrophils were examined microscopically for the presence and absence of cell-associated bacteria. The number of bacteria per cell was calculated for those leucocytes which appeared to have ingested bacteria.
Assay of Phagocytosis Using Differential Centrifugation In these experiments, leucocyte/bacteria suspensions After identical periods of were prepared as described. incubation, 0’2 ml. of the mixture was removed and washed twice in H.B.S.S. to remove the extracellular bacteria. The cell pellet was lysed in distilled water and the viable cell-associated bacteria were quantitated as described. No antibiotics were added to the assay.
Results
Phagocytic Capacity ofLeucocytes The phagocytic capacity of leucocytes from four patients with C.G.D., from two known female carriers of C.G.D., and from controls was assessed using lysostaphin (fig. 1). In the experiments shown, 5-0 X 101 neutrophils were tested with 106 bacteria and aliquots were examined after 5, 10, and 20 minutes of incubation. As shown in fig. 1, the capacity of the patients’ leucocytes to phagocytise Staph. aureus exceeded the phagocytic capacity of control leucocytes. In experiments not shown here, similar results were obtained when the bacteria/leucocyte ratio was varied by
Preparation of the Test Organisms Staph. aureus 502A and Streptococcus faecalis were used as test organisms after an 18-hour culture. Just prior to use, bacteria were washed three times in physiological saline solution and resuspended in physiological saline to contain approximately 107 organisms per ml. The precise number of bacteria was assessed for each experiment using standard dilution and pour-plate techniques.
Assay of Phagocytosis Using Lysostaphin The method used
examine the kinetics of early phagocytosis by leucocytes was modified from the method described by Tan et al.1o Briefly, 0-5 ml. of the leucocyte suspension, 0’1 ml. pooled normal human serum (P.N.H.s.), 0’1 ml. bacterial suspension, and 0’3 ml. H.B.s.s. were added to a sterile 12 X 75 mm. disposable Falcon tube and the mixture was incubated at 37°C on an aliquot mixer (Lab-Tek, Inc., Westmont, Illinois). Control tubes, one without leucocytes (bacteria control) and the other without serum, were examined in parallel. After 5, 10,’ and 20 minutes of incubation, a 0-2 ml. aliquot was removed from each tube. To each aliquot was added 3’5 units of lysostaphin (Mead Johnson Research Center, Evansville, Ind.), and the mixture was incubated at 37°C for 20 minutes. To inactivate the lysostaphin, 0-015 ml. of 2-5% trypsin isotonic saline solution was added and the mixture was incubated for 10 minutes. The incubation mixture was then diluted with 1’75 ml. sterile distilled water to lyse the leucocytes, and the number of viable intracellular bacteria was quantitated using standard dilution and pour-plate techniques. In several experiments, different bacteria to neutrophil ratios were examined (1/5 to 100/1). Since several drug-induced abnormalities of leucocyte function have features similar to the leucocyte abnormalities described in C.G.D.,14,15 the kinetics of phagocytosis was assessed in the presence of 2’5xlO’M phenylbutazone. to
phagocytosis of Staph. aureus by leucocytes patients with C.G.D. (.———A), carriers of C.G.D. (D———D). and controls (0-0), using lysostaphin.
Fig. 1-Early from
The IOgl. number viable bacteria per ml. is plotted versus time in minutes. Data represent the means.E. of eleveo experiments for patients and controls and the range of two experiments for two carriers. Rate of killing Staph. aureus by lysostaphin ( - - - - W) and bacteria control (N - - - -N).
993
and the variability of the assay. While this technique was less sensitive than the lysostaphin method, which
neutralised extracellular Staph. aureus, differential centrifugation could detect differences, although much smaller, in early phagocytosis of Staph. aureus between leucocytes from C.G.D. patients and controls. In separate experiments in this laboratory (data not shown) and by Holmes and GOOd,14 leucocytes from C.G.D. patients killed Strep. f (zcalis
rapidly
normally.
,
Influence of Phenylbutazone on Phagocytosis The phagocytic capacity of normal leucocytes for Stap7i. aureus in the presence of 2.5 X 10--’M phenyl-
Fig. 2-Representative experiment comparing early phagocytosis of Strep. feecalis by leucocytes from patients with C.G.D. (A.——A ), carriers of C.G.D. ( D—— D): and controls ( 0—— 0). Using differential centrifugation, the loglo number viable bacteria is plotted versus time in minutes.
butazone was increased 5-10 fold when tested in four experiments. By contrast, phenylbutazone had no effect on the phagocytic capacity of C.G.D. leucocytes. In parallel experiments (data not shown) and in studies by Holmes et aL14 similar concentrations of this drug induced a bactericidal abnormality in normal
leucocytes similar
to
that in C.G.D.
leucocytes.
Discussion
10-fold increments from 1 bacterium per leucocyte An increased phagoto 100 bacteria per leucocyte. for aureus Staph. cytic capacity by C.G.D. leucocytes was also suggested in experiments using differential centrifugation techniques, although the differences were much smaller. Further, when smears of leucocytes were assessed morphologically for phagocytosis after 10 and 20 minutes’ incubation with Staph. aureus, the number of bacteria associated with C.G.D. leucocytes appeared to be 2 to 5 fold greater than the number of bacteria associated with control leucocytes. It is also apparent in fig. 1 that phagocytosis by the patients’ leucocytes was maximum after only 5 minutes of incubation and did not appear to increase significantly with longer incubation. By contrast, with longer periods of incubation normal leucocytes continued to phagocytise bacteria, and at the times studied their phagocytic capacity did not equal the phagocytic capacity of C.G.D. leucocytes. In two experiments, leucocytes from two carriers of C.G.D. appeared to be intermediate in their capacity to phagocytise Staph. aureus. Patients’ sera did not influence the phagocytic capacity of normal leucocytes. Under the experimental conditions employed, no significant bacterial growth occurred for up to 3 hours of observation (fig. 1). A
representative experiment comparing the phagocytic capacity of leucocytes from the patients and controls for Strep. fcccalis is summarised in fig. 2. In these experiments, the phagocytic capacity of
leucocytes was assessed after similar times of incubation using differential centrifugation to separate the cell-associated bacteria from non-cell-associated, or supernatant, bacteria. As shown in fig. 2, phagocytosis of Strep. facalis by C.G.D. leucocytes was similar to that by control leucocytes. In four experiments, the rate of phagocytosis of Strep. facalis by leucocytes was less than with control leucocytes. This difference, although seemingly reproducible, may not be significant when one considers the sensitivity C.G.D.
Impaired antimicrobial activity of leucocytes can be due to abnormalities in the ability of leucocytes to ingest bacteria or to an intracellular microbicidal defect.16-18 Furthermore, microorganisms themselves appear to influence, in a major way, the extracellular and the intracellular phases of microbial inactivation.19 diminished bactericidal and metabolic releucocytes from patients with C.G.D. have well documented.’,’By contrast to the sophisticated metabolic studies, the methods of examining the kinetics of phagocytosis by leucocytes have been much less precise. Thus, the phagocytic capacity of leucocytes from C.G.D. patients, although generally reported to be normal or slightly diminished, has never been examined with a sensitive The
sponses of
tbeen
technique.3 The observations reported here demonstrate that the phagocytic capacity of leucocytes from patients with C.G.D. varies considerably with the test organisms and the assays used. Phagocytosis by leucocytes from patients with C.G.D. of Staph. aureus, an organism not readily killed by these cells, appears to be greater than phagocytosis by control leucocytes (fig. 1). Indeed, leucocytes from patients with C.G.D. phagocytised as many as 300-fold the number of Staph. aureus organisms which were phagocytised by control leucocytes. In these experiments, the time required by leucocytes from patients with C.G.D. to achieve maximum phagocytosis appeared to be very short. Leucocytes from patients with C.G.D., when tested at several different bacteria/leucocyte ratios, achieved maximum phagocytosis after only 5 to 10 minutes’ incubation. By contrast, the rate of early phagocytosis by normal leucocytes appeared to be slower. Quantitation of phagocytosis by normal leucocytes after periods of incubation greater than 20 minutes was difficult to assess since, after 20 minutes of incubation, a significant number of intracellular bacteria were being killed by the leucocytes. Many factors appear to influence the union of phagocytic cells with bacteria, both in vitro and
994 factors have been retard this process. can enhance or Indeed, a fragment of gamma-globulin has been reported to promote the phagocytosis of staphylococci by phagocytic cells z° A serum factor would not appear to account for the observations reported here, since P.N.H.S. and sera from patients with C.G.D. supported phagocytosis of control leucocytes and patient leucocytes equally. In addition, significant intracellular multiplication of bacteria in C.G.D. leucocytes would not account for the observations reported here since multiplication in the bacterial controls in each experiment was negligible (fig. 1). In order to examine these observations further, the phagocytic capacity of leucocytes from patients with C.G.D. was assessed using a test microorganism which their leucocytes could normally kill. These studies were important in order to determine if the phagocytic capacity of C.G.D. leucocytes for Staph. aureus was enhanced nonspecifically, or was influenced perhaps in some way by the ability of the leucocyte to kill the test microorganism. As shown in fig. 2, using a less sensitive assay of phagocytosis Strep. fœcalis, an organism which was killed vigorously by C.G.D. leucocytes,14. appeared to be ingested as readily by control leucocytes as by C.G.D. leucocytes. One possible explanation of these observations is that the phagocytic capacity of a leucocyte is influenced, in some way, by its ability to kill an ingested microorganism. Thus, the superior rate of early phagocytosis of Staph. aureus by C.G.D. leucocytes may represent an attempt by the patients’ leucocytes to compensate for their abnormal bactericidal capacity. Some drug-induced abnormalities of leucocyte function are similar to the bactericidal abnormality of C.G.D. leucocytes. 14 Although drugs are seldom selective in their influence on leucocyte metabolism, they have proven to be promising laboratory models of C.G.D. Phenylbutazone inhibits the bactericidal capacity of leucocytes for both catalasepositive and catalase-negative bacteria,14,15 but has no significant effect on phagocytosis. This laboratory model was then utilised to examine the possible effects of this drug on the early phagocytic capacity of normal leucocytes for Staph. aureus. In four experiments, phenylbutazone appeared to enhance the phagocytic capacity of normal leucocytes for Staph. aureus. This observation can be explained, in part, by the effect of phenylbutazone on bacterial killing by neutrophils. The bacteriostatic and bactericidal effects of neutrophils on ingested bacteria are expressed soon after the bacteria have been ingested and vary with the type of bacteria ingested. Although the bacteria may be unable to multiply, they may remain relatively intact and capable of continued cellular metabolism for longer periods.19 Active degradation of bacteria by leucocytes requires more time and again may vary considerably with the type of microorganism ingested. For staphylococci, with their inherent ability to resist both phagocytosis and killing by leucocytes, the leucocyte begins to digest the ingested microorganisms after approximately an hour.17 Differences between the bactericidal capacity of normal leucocytes and of C.G.D. leucocytes for Staph. aureus probably account for some of the observations reported in vivo.
In
vitro, several
described which
’""
serum
here. Normal leucocytes, by their early bacteriostatic effect on ingested bacteria, would have fewer viable intracellular bacteria, and would therefore appear to have phagocytised fewer bacteria than C.G.D. leucocytes which do not exert this bacteriostasis. This explanation, however, would not appear to account for the large difference in phagocytosis reported here since phagocytosis by leucocytes from patients with C.G.D., when assessed morphologically and by differential centrifugation, appeared to exceed that of control
leucocytes. Leucocytes
from female carriers of C.G.D. are intermediate between normal and leucocytes from patients with C.G.D. in their capacity to kill Staph. aureus in vitro. Approximately half of their peripheralblood leucocytes appear to have a metabolic defect similar to the C.G.D. patients’ 12,13 and half appear normal. However, this division between normal and abnormal cells may be quite variable as is illustrated by a female carrier described previously with bactericidal and metabolic abnormalities very similar to those in her brother with C.G.D.13 Despite these abnormalities of leucocyte function, which, in some carriers may be considerable,13 female carriers do not appear unduly susceptible to bacterial infection. While many more patients and carriers need to be studied, the two carriers studied in this report had an increased capacity to phagocytise Staph. aureus (fig. 1). This increased phagocytic capacity may function as an important compensatory mechanism by leucocytes in carriers of C.G.D. Such a mechanism in carriers might circumvent an increased susceptibility to bacterial infection and at least partly account for their clinical well-being. While more carriers of C.G.D. need to be studied, this assay of phagocytosis using lysostaphin appears to be useful for detecting C.G.D. carriers and to be relatively simple compared to previously described methods for the detection of C.G.D. carriers. These studies seem to demonstrate another aspect of leucocyte function and may provide some information on the recognition and controlling mechanisms of leucocytes. Clearly, greater understanding of the phagocyte cell-membrane and of the physical and biochemical properties of the surface of the microorganisms that influence particle ingestion is needed. I thank K. Y. Wong for her excellent technical assistance and Dr A. R. C. Cole and Dr P. D. McClure for permission to study their patients. This study was supported by the Medical Research Council of Canada (Grant MA5276). W. D. B. is a Medical Research Council scholar. Requests for reprints should be sent to W. D. B., Department of Immunology, Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada M5G 1X8. REFERENCES 1. Holmes, B., Page, A. R., Good, R. A. J. clin. Invest. 1967, 46, 1422. 2. Baehner, R. L., Nathan, D. G. Science, 1967, 155, 835. 3. Karnovsky, M. L. Fedn Proc. 1973, 32, 1527. 4. Holmes, B., Park, B. H., Biggar, W. D. in Molecular Pathology (edited by C. C. Thomas and S. B. Day) (in the press). 5. Klebanoff, S. J., Pincus, S. H. J. clin. Invest. 1971, 50, 2226. 6. Mandell, G. L., Hook, E. W. J. Bact. 1969, 100, 531. 7. Stossel, T. P., Mason, R. J., Hartwig, J., Vaughan, M. J. clin. Invest.
1972, 51, 615. 8. Schindler, C. A., Schuhardt, V. T. Biochem. 97, 242.
Biophys. Acta, 1965,
995
LYMPH-NODE BIOPSY DURING SIMPLE MASTECTOMY A. A. SHIVAS ELIZABETH L. M. CANT A. P. M. FORREST
Departments of Clinical Surgery and Pathology,
University of Edinburgh The distribution of pectoral (external mammary) nodes identified during the removed with the axillary tail of the and operation breast was studied in 45 patients treated by simple (total) mastectomy. Up to 13 nodes may lie within the axillary tail, and these are continuous with the pectoral nodes. Lymph-nodes were identified in 90% of patients treated by simple (total) mastectomy without dissection of the axilla. Sum ary
tail postoperatively; and "pectoral nodes" as those selected peroperatively by the surgeon (fig. 1). Methods
Simple mastectomy was performed in 45 patients with early breast cancer. Our technique for this procedure is described elsewhere.5 It includes careful definition and dissection of the axillary tail of the breast from between the pectoralis major muscle in front and the latissimus dorsi behind, and its removal with the breast up to the point where it blends with the axillary fat-i.e., at the level of the third rib. This is facilitated by removing the
Introduction WE have suggested that a rational policy for the local management of primary breast cancer is simple mastectomy with pectoral-node biopsy, followed by immediate postoperative radiotherapy only in those cases in which involvement of these nodes is proved histologically. Since these nodes can be identified without dissecting the axilla, the morbidity of either unnecessary axillary dissection or radical radiotherapy is avoided. We reported a controlled randomised study which was designed to assess this policy of treatment in Cardiff and St. Mary’s Hospital.’ This is now also the basic procedure used in the current Edinburgh breast cancer trials, which include the administration of additional systemic therapy to those with histologically involved pectoral lymph-nodes.3.4t We have been uncertain as to the true distribution of these pectoral nodes. Initially we believed that be near to the top of the best identified they could excised breast specimen, but later axillary tail of the we advised that the surgeon should identify a node or nodes of this group during his dissection of the axillary tail, particularly at the point of separation from the axillary fat. The study we now report was undertaken to determine the best method of accurate sampling of these nodes. It suggests that a careful search of the excised axillary tail and a careful search by the surgeon peroperatively are both essential steps for complete sampling. For clarity, we have defined " axillary-tail nodes " as those identified in the axillary
9. Schaffner, W., Melly, M. A., Hash, J. H., Koenig, M. G. J. biol. Med. 1967, 39, 215. 10. Tan, J. S., Watanakunakorn, C., Phair, J. P. J. Lab. clin. Med. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.
1971, 78, 316. Holmes, B., Quie, P. G., Windhorst, D. B., Good, R. A. Lancet, 1966, i, 1225. Windhorst, D. B., Page, A. R., Holmes, B., Quie, P. G., Good, R. A. J. clin. Invest. 1967, 47, 1026. Biggar, W. D., Buron, S., Holmes, B. J. Pediat. (in the press). Holmes, B., Good, R. A. J. reticul. Soc. 1972, 12, 216. Strauss, R. R., Paul, B. B., Sbarra, A. J. J. Bact. 1968, 96, 1982. Stossel, T. P. New Engl. J. Med. 1974, 290, 717. Quie, P. G. in Current Problems in Pediatrics, vol. XI, no. 11, p. 1. Chicago, 1972. Baehner, R. L. J. Pediat. 1974, 84, 317. Elsbach, P. New Engl. J. Med. 1973, 289, 846. Najjar, V. A., Constantopoulos, A. J. reticul. Soc. 1972, 12, 197.
Fig. 1-Position of axillary-tail and pectoral lymph-nodes. breast from the medial to lateral side. Separation of the tail from the axillary fat pad is facilitated by grasping it and feeling between finger and thumb the point at which the nodular breast fat merges into the much finer and smoother axillary fat lobules. After removal of the breast, the axillary tail was detached and the cut end marked with a suture. It was spread on a cork board and palpated on the flat. Palpable nodes were removed and their site marked on a sketch plan. The axillary tail was then sliced and any further nodes removed and also marked on the plan. Each node was fixed in 10% formol saline solution, embedded in paraffin, and sectioned. These nodes are termed " axillary-tail nodes ". Several sections were taken from each half of the node and stained with haematoxylin and eosin. When examination of these initial sections was negative, serial 1/10 sections were cut from the nodes, stained, and examined for foci of tumour. In all cases, sections were taken to confirm the presence of breast parenchyma in the axillarytail specimens. In 38 cases, at operation the surgeon also attempted to identify a node or nodes separate from the axillary tail These usually lay at the junction of the axillary tail and the axillary fat, and towards its medial aspect. A node was defined in 29 cases; 2 of these on section proved to be only fat. These nodes have been termed " pectoral
Results
Total Yield of Nodes from Axillary Tail Nodes were identified within the breast of the axillary tail in 32 (71 %) of the 45
parenchyma cases.
The
