Discrepancies in Fluorescent Antibody, Counterimmunoelectrophoresis, and Neufeld Test for Typing of Streptococcus pneumoniae

509

DIAGN MICROBIOL INFECT DIS 1985;3:509-514

Discrepancies in Fluorescent Antibody, Counterimmunoelectrophoresis, and Neufeld Test for Typing of Streptococcus pneumoniae Ashir Kumar, Jose Mariappuram, and Chang H. Kim

The techniques of fluorescent antibody (FA) and counterimmunoelectrophoresis (CIE) were compared with the Neufeld test (quellung reaction) for typing of Streptococcus pneumoniae. A total of 88 isolates were examined by these three methods. Pool-, type-, or group-specific pneumococcal antisera were used in all three methods. Each isolate was initially tested with polyvalent antisera and all of the nine pools of antisera. Selection of the type- or group-specific antisera depended upon the reaction of the isolate with. the pool sera. Sixty-eight of 88 (77.3%) isolates were positive using pool or typing sera and were correctly typed by CIE, while FA was found to be accurate for only 61 of 88 (69.3%) isolates. Positive reactions with more than one pool- or type-specific antisera, or no reaction, were seen with several of the isolates with both techniques. Even though CIE and FA are rapid and simple techniques, microbiologists should be cautious when utilizing them for typing of S. pneumoniae because of the discrepancies observed in this study.

INTRODUCTION Over the last decade several new techniques for rapid identification of organisms have been employed in microbiology laboratories. Coonrod and Rytel (1973) described the specificity of counterimmunoelectrophoresis (CIE) in identifying typespecific pneumococcal antigens. Since then CIE has achieved wide acceptance for the detection of antigen in various body fluids and in vitro experiments (Artman et al., 1980; Corgeni and Nankervis, 1978; Kumar and Kumar, 1981; Naiman and A1britten, 1980; Waters et al., 1979). Wicher et al. (1981) compared the efficacy of fluorescent antibody (FA) technique for identifying and typing of Streptococcus pneumoniae with the classic time-proven Neufeld test. In this study we utilized CIE and FA techniques for typing and identification of S. pneumoniae and compared the results with the Neufeld test. MATERIALS AND METHODS Streptococcus pneumoniae isolates from sputum, throat culture, blood, or spinal fluids were utilized. The isolates were grown on Trypticase soy agar with 5% sheep erythrocytes at 37°C in a CO2 jar and were identified by colony appearance, Gram stain, and optochin susceptibility.

From the Department of Pediatrics, Saint Lukes Hospital, Cleveland, Ohio (J.M., C.H.K.) and the Department of Pediatrics and Human Development, Michigan State University, East Lansing, Michigan (A.K.). Address reprint requests to: Ashir Kumar, M.D., Department of Pediatrics and Human Development, B240 Life Sciences Building, East Lansing, MI 48824. Received October 15, 1984; revised and accepted January 21, 1985. © 1985 Elsevier Science Publishing Co., Inc. 52 Vanderbilt Avenue, New York, NY 10017

0732-8893/85/$03.30

510

A. Kumar et al.

Diagnostic Antisera Pneumococcal antisera obtained from Statens Serum Institute (Copenhagen, Denmark) were utilized. These sera are p r o d u c e d in rabbits and included: 1) o m n i s e r u m , a polyvalent a n t i s e r u m containing antibodies to all 83 k n o w n p n e u m o c o c c a l capsular polysaccharide-type antigens; 2) nine p o o l e d sera, designated A to I, each reacting with 7-11 types of antigens a n d together covering all 83 types; and 3) t y p e or group sera reacting w i t h single types or groups n u m b e r e d 1 to 48.

Neufeld Test The capsule reaction test or quellung reaction was performed using the standard technique: a small drop (0.5 ml) of broth culture was placed on a glass slide, 0.05 ml of antiserum was dropped, and 0.05 ml of 0.2% methylene blue was a d d e d to this; the reagents were gently mixed, a cover slip was put over the mixture, and the capsular swelling was evaluated using an o i l - i m m e r s i o n magnification a n d decreased illumination. A l l isolates were first tested w i t h o m n i s e r u m and later w i t h each of nine pooled sera. Once the isolate was classified to a pool, then only type- or groupsspecific sera belonging to that particular pool were used for the quellung reaction.

Countercurrent Immunoelectrophoresis The CIE m e t h o d d e s c r i b e d previously was e m p l o y e d (Kumar and Kumar, 1981). The CIE was carried out on 1 × 3 inch glass slides; the slides were covered w i t h 3.0 ml of 1% agarose in Veronal buffer with a p H of 6.6 and 0.05 molar. Veronal buffer of pH 8.2 was used in the chamber. Parallel rows of eight wells, 3 m m in diameter, were cut 5 m m apart in the agar. Isolated strains of S. pneumoniae to be tested were grown in trypticase soy broth for 24 hr; 15-20 ~,1 of this broth culture filled the wells on the cathode side. O m n i s e r u m and other p o o l e d antisera (designated A to I) filled the wells on the a n o d e side. Each isolate was initially tested w i t h p o l y v a l e n t omn i s e r u m and all n i n e p o o l antisera. Selection of the t y p e or group sera d e p e n d e d u p o n the pool to w h i c h an isolate reacted. If an isolate gave p r e c i p i t i n lines w i t h more t h a n one p o o l antisera, then type or group sera covering those pools were included. The slide was connected to the buffer in the chamber w i t h a moist gauze. Electrophoresis was carried out at a r o o m t e m p e r a t u r e using 11 m A p e r slide 3 hr. Slides were read u n d e r indirect light to d e t e r m i n e if precipitin lines had formed. All slides were t h e n refrigerated overnight in a moist chamber and read again for precipitin lines.

Fluorescent Antibody Technique

Streptococcus pneumoniae to be tested were grown in trypticase soy broth for 24 hr. The tubes were centrifuged at 3000 RPM for 10 min and the pellets were w a s h e d twice with phosphate-buffered solution (PBS) pH 7.4 to 7.8. The pellets were resusp e n d e d in a few d r o p s of PBS, and films of this suspension were p r e p a r e d on glass slides, air dried, a n d fixed in 95% ethanol for 10 min. These films were covered w i t h 0.1 ml diluted p n e u m o c o c c a l antisera in a moist chamber at room t e m p e r a t u r e for 30 min. All isolates were tested with o m n i s e r u m and pooled antisera, however, an isolate was tested w i t h only those types or groups of antisera that belonged in the p~rticular pool to w h i c h the isolate gave a positive fluorescent reaction. The slides were then w a s h e d w i t h PBS for 15 min. F l u o r o e s c i n isothiocynate congugated goat a n t i r a b b i t g l o b u l i n (obtained from Cappel Laboratories) diluted 1 : 3 2 was p r o p e r l y placed onto slides a n d the slides were again incubated at room temperature in the

Discrepancies i n T y p i n g S.

pneumoniae

511

dark for 30 m i n in a moist chamber. Slides were washed in PBS twice a n d air dried. The slides w e r e m o u n t e d with glycerol and examined u n d e r fluorescent microscope immediately for the presence of fluorescence. Most bacterial cells were able to retain the fluorescence for 24 hr if they were kept i n a dark chamber in the refrigerator.

Analysis of Data Reaction of each isolate to omnisera, pool sera, and type or group sara by each of the three methods was noted. The results of CIE and FA were compared with the identification obtained by the Neufeld test.

RESULTS A total of 88 isolates of S. pneumoniae were examined by the three methods. All isolates reacted with omnisera, one pool antisera, and one type- or group-specific antisera during the Neufeld test. Each isolate demonstrated quellung only with one pool antisera, and then reacted with only one type- or group-specific antisera w i t h i n that pool. Type 19 was the c o m m o n e s t type identified (21/88, 23%) followed by type 3 (18/88, 20%), type 23 (15/88, 17%), and type 6 (8/88, 9%). Type and pool types of all isolates tested are given in Table 1. Sixty-eight of 88 (77.3%) could be correctly typed by u s i n g CIE, w h i l e FA was found to be accurate in only 61 of 88 (69.3%) of the isolates. Three types of identification discrepancies were observed with CIE a n d FA techniques: 1) no reaction to type-specific and pool antisera; 2) no reaction with type-specific antisera only; and 3) false positive reactions with type-specific and/or pool antisera (Table 2). False reactions or no reactions by CIE and FA techniques were not necessarily seen i n the same isolate, thus some isolates were identified correctly by CIE and not by FA a n d vice versa.

TABLE 1. Correct Identification of S. p n e u m o n i a e Isolates (Type and Pool) Using CIE and FA T e c h n i q u e s Total isolates

Correctly identified

(Neufeld test) Type

Pool

1 3 6 8 9 10 11 13 14 15 18 19 23 29

A B B B D E D H H H A B H G

Totals

CIE

FA

Isolates (n)

(n)

(%)

(n)

(%)

5 18 8 2 1 2 3 1 4 5 1 21 15 2

5 12 7 1 O O 3 1 0 5 I 17 14 2

100 66.7 87.5 50 0 0 100 100 0 100 100 80.9 93.3 100

4 9 6 1 0 0 3 1 O 5 1 15 14 2

80 50 75 50 0 0 100 100 0 100 100 71.4 93.3 100

68

77.3

61

88 (100%)

69.3%

512

A. Kumar et al.

TABLE 2. Identification Discrepancies Observed with CIE and FA Techniques

Discrepancies A. Isolates w i t h n o r e a c t i o n to type a n d pool a n t i s e r a

C1E t e c h n i q u e

FA technique

T y p e - p o o l (n)

T y p e - p o o l (n)

3-B 8-B 9-D 10-E 14-H 19-B

Subtotal B. Isolates w i t h no r e a c t i o n to type-specific a n t i s e r a

3 1 1 2 3 1

1-A 3-B 6-B 8-B 9-D lo-E 14-H 19-B

1 7 2 1 1 2 3 4 21

3-B 14-H 23-H

11

Subtotal

2

2 1 1 4

C. IsoLates w i t h false identification Subtotal

7a

2a

3-B

Total d i s c r e p a n c i e s

2

20

27

°Details of the isolate's type and pool are given in the text.

Discrepancies Observed Using CIE for Typing Eleven isolates failed to demonstrate any precipitin lines with type and pool antisera. Two isolates of type 3 demonstrated precipitin lines with antisera for pool B but could not be further typed. Seven of 88 (8%) isolates gave false identifications with CIE technique: One type 3 isolate was identified as type 6, one type 14 isolate was identified as type 19, one type 23 isolate was identified as type 15, one type 19 isolate demonstrated precipitin lines with both type 6 and type 19 antisera. One pool B isolate also had precipitin lines with both pool B and pool H antisera, one type 19 (pool B) isolate reacted with pool D antisera, and one type 19 (belonging to pool B) isolate demonstrated precipitin lines with both pool I and pool H antisera but not with pool B antisera. A total of 20 identification discrepancies were observed with the CIE technique. These discrepancies were not linked to certain specific types but were distributed among nine of the 14 types studied. Six of 18 (33%) type 3 isolates could not be serotyped by CIE, one of them gave precipitin line with type 6 antisera, five of them did not form precipitins with any of the type-specific antisera. Three of the type 3 isolates could not even be correctly identified in pool B by CIE. Few isolates of type 8, type 9, and type 10 could not be correctly identified to their particular pool or type with this method. None of the four isolates of type 14 could be identified by CIE. One type 14 isolate formed precipitin lines with pool B and type 19 antisera. Four of 21 type 19 isolates gave false positives or no reactions when CIE was utilized; some of these isolates formed precipitin lines with pool D, pool H, pool I and type 6 antisera.

Discrepancies in Typing S. pneumoniae

513

Discrepancies Observed Using FA Method for Typing Twenty-one of 88 (24%) isolates did not demonstrate any fluorescence with the FA method and thus could not be identified in a pool or further typed (Table 2). Four isolates (two of type 3, one each of types 14 and 23) could be correctly pooled but could not be further typed utilizing this technique. Only 2 of 88 (2.3%) isolates were falsely typed with the FA technique. Both these isolates were type 19 by the Neufeld method, but gave false-positive fluorescence with type 6 antisera. A total of 27 identification discrepancies were observed with the FA technique. These discrepancies were not linked to certain type-specific isolates but were distributed among 9 of the 14 types studied. Nine of 18 (50%) type 3 isolates could not be serotyped by the FA method. Seven isolates identified as type 3 by the Neufeld test had a negative FA test with pool B and type 3 antisera; two additional isolates had a negative FA test to only type 3 antisera but demonstrated fluorescence with pool B antisera. Two isolates of type 6 did not demonstrate any fluorescence with type 6 and pool B antisera. Few isolates of type 8, type 9, and type 10 could not be correctly identified to their particular pool or typed by this technique. None of the four isolates of type 14 could be typed by FA. One type 4 isolate was correctly identified as belonging to pool H but could not be further typed, another three isolates did not react at all in the FA technique. Six of 21 (29%) type 19 isolates gave false positives or no reactions when FA was utilized; four of these did not react by the FA method, and another type 19 isolate demonstrated false positivity with type 6 antisera.

DISCUSSION Approximately 23% and 31%, respectively, of the isolates of S. pneumoniae could not be correctly typed by CIE and FA techniques in this study. Nine of 88 (10.2%) isolates gave false-positive reactions with type and pool antisera and thus would have been misidentified if only CIE was used. Six of 88 (6.8%) isolates gave falsepositive reactions with type- and pool-specific antisera with the FA technique and thus would have been mistyped if only the FA technique was utilized. Wicher et al. (1981) described FA to be a useful and practical method for identifying as well as typing S. pneumoniae and they also pointed out the cross-reactions that are often seen with other microorganisms when the FA technique is used. In this study we found cross-reactions or no reaction with type- and pool-specific antisera when CIE and FA were employed. The cross-reactivity and no reactions were seen with several pneumococcal types. Coonrod and Rytel (1973) pointed to the relative lack of sensitivity of CIE for detection of types 7 and 14 pneumococcal antigens. None of the type 14 pneumococci demonstrated a precipitin with type 14 or pool H antisera; one type 14 reacted with type 19 and pool B antisera. Type 7 and type 14 pneumococci are not routinely detected by CIE because their antigens do not carry the proper electrical charge and migrate in the wrong direction when electrophoresis is carried out (Anhait and Yu, 1975). We were unable to type any one of the four S. pneumoniae type 14 isolates with the FA technique. One of the isolates did react with pool H antisera in FA techniques. Type 14 pneumococcus has repeatedly been shown to be one of the common types that causes pneumonia in children; the fact that this type could not be identified by the rapid methods of CIE and FA represents a major drawback of these techniques (Finland, 1973). Wicher et al. (1981) found FA to be superior to Gram stain and cultures when looking for pneumococci in direct films of various body fluids. Several investigators have found the technique of CIE to be very valuable in the identification of pneumococcal antigen in various body fluids (Artman et al., 1980; Congeni and Nankervis, 1978; Coonrad and Rytel, 1973; Kumar and Kumar,

A. K u m a r et al.

514

1981; N a i m a n a n d Albritton, 1980; Waters et al., 1979). In most of those situations omniserum is u s e d a n d the p o o l and t y p e specificity of a particular p n e u m o c o c c u s is of little significance. Cross-reactivity among different types of S. p n e u m o n i a e cannot be avoided, a n d the Statens Serum Institute informs investigators regarding the cross-reactivity of these antisera. However, that does not e x p l a i n w h y crossreactivity was so frequently observed in this s t u d y with the CIE t e c h n i q u e and not with the quellung reactions or the FA techniques; it also does not e x p l a i n w h y certain isolates of S. p n e u m o n i a e did not react w i t h their own type-specific a n d pool antisera but gave false-positive reactions with other antisera in the CIE and F A methods. Typing of S. p n e u m o n i a e is not routinely carried out in most hospital laboratories as it does not have m u c h bearing on patient care. However, w h e n p n e u m o c o c c a l infections are being s t u d i e d from an e p i d e m i o l o g i c or infection control p o i n t of view, knowing the type of a particular isolate is of significance. This s t u d y has p o i n t e d out the discrepancies that are possible if only CIE or F A techniques are used for typing. This study was supported by a research grant from the Board of Trustees of Saint Lukes Hospital, Cleveland, Ohio. REFERENCES

Anhalt JP, Yu PKW (1975) Counterimmunoelectrophoresis of pneumococcal antigens: Improved sensitivity for the detection of types VII and XIV. J Clin Microbiol 2:510. Artman M, Weiner M, Frankl G (1980) Counterimmunoelectrophoresis for early detection and rapid identification of Haemophilus influenzae type b and Streptococcus pneumoniae in blood cultures. J Clin Microbial 12:614. Congeni BL, Nankervis GA (1978) Diagnosis of pneumonia by counterimmunoelectrophoresis of respiratory secretions. Am J Dis Child 132:684. Coonrod JD, Rytel MW (1973) Detection of type-specific pneumococcal antigens by counterimmunoelectrophoresis. 1. Methodology and immunologic properties of pneumococcal antigens. J Lab Clin Med 81:770. Finland M (1973) Excursions into epidemiology: Selected studies during the past four decades at Boston City Hospital. J Infect Dis 128:76. Kumar A, Kumar K (1981) Rapid laboratory diagnosis of infectious diseases. Primary Care 8:593. Naiman HL, Albritton WL (1980) Counterimmunoelectrophoresis in the diagnosis of acute infection. J Infect Dis 142:524. Waters C, Fossieck B, Parker RH (1979) Counterimmunoelectrophoresis for rapid identification of blood-culture isolates. Am l Clin Path 71:330. Wicher K, Kalinka C, Mlodozeniec P, Rose NR (1981) Fluorescent antibody technique used for identification and typing of streptococcus pneumoniae. Am J Clin Pathol 77:72.