Detection of serum immunoglobulin M to human cytomegalovirus by ...

CLINICAL AND DIAGNOSTIC LABORATORY IMMUNOLOGY, Sept. 1996, p. 597–600 1071-412X/96/$04.0010 Copyright q 1996, American Society for Microbiology

Vol. 3, No. 5

Detection of Serum Immunoglobulin M to Human Cytomegalovirus by Western Blotting Correlates Better with Virological Data than Detection by Conventional Enzyme Immunoassay T. LAZZAROTTO,1 G. T. MAINE,2 P. DAL MONTE,1 H. FRUSH,2 K. SHI,2

AND

M. P. LANDINI1*

Department of Clinical and Experimental Medicine, Section of Microbiology, University of Bologna, Bologna, Italy,1 and Abbott Laboratories, Abbott Park, Illinois2 Received 26 February 1996/Returned for modification 6 May 1996/Accepted 29 May 1996

Western blotting (immunoblotting) with proteins separated from purified human cytomegalovirus (HCMV) particles (viral WB) has repeatedly been shown to be a reliable and sensitive method for detecting HCMVspecific immunoglobulin M (IgM). The aim of the present work was to determine whether IgM detected by viral WB correlates with virological diagnosis better than conventional enzyme immunoassay (conv-EIA). The presence of an active HCMV infection was documented on the basis of isolation of virus from urine and/or saliva and on the basis of antigenemia and/or PCR with polymorphonuclear leukocytes for immunocompetent and immunocompromised subjects, respectively. The agreement observed between IgM detected by viral WB and the results obtained by virological detection of HCMV was significantly higher (88.7%) than the agreement of IgM detected by conv-EIA and virological results (67.5%). of the presence of the virus or viral components (mainly the viral genome and viral antigens) in pathological materials, continues to be the reference procedure for the determination of an active HCMV infection (2, 8, 10, 16, 22). The aim of the present work was to determine whether IgM detected by viral WB correlates with virological diagnosis when the virological diagnosis is determined by virus isolation and by the relatively new procedures of detection such as antigenemia and PCR. A total of 717 serum samples were included in this study. Of these, 210 serum specimens were from immunocompetent subjects (35 infants, 64 pregnant women, and 111 blood donors) while 507 were from immunocompromised patients (228 renal transplant recipients, 156 heart transplant recipients, 42 liver transplant recipients, 58 bone marrow transplant recipients, and 23 patients suffering from AIDS). For immunocompetent subjects, the presence of an active HCMV infection was documented on the basis of isolation of virus from urine and/or saliva. For immunocompromised patients, the presence of an active HCMV infection was determined by antigenemia and/or PCR with polymorphonuclear leukocytes (PMNL). The Towne strain of HCMV was propagated in human embryo fibroblasts by standard methods. The virus was purified by sorbitol gradient, and viral proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electrotransferred onto nitrocellulose as previously described (13). The evaluation of anti-HCMV IgM was carried out in parallel with conventional EIA (conv-EIA) and viral WB. For conv-EIA, the Enzygnost anti-HCMV/IgM kit (Behring AG, Marburg, Germany) was used. The assay was performed as suggested by the manufacturer, and results for the plates were determined with a microEIA automatic reader (Behring AG). The detection of anti HCMV-IgM by viral WB was done as previously described in detail (13). The viral WB was interpreted as follows. A serum specimen was considered positive by WB if an antibody reaction was observed with one or more viral proteins with a molecular mass of 150, 65, 55, or 38 kDa. A serum specimen was considered negative by WB if no anti-

Human cytomegalovirus (HCMV) is responsible for severe and often life-threatening infections in immunocompromised patients, such as transplant recipients and patients with AIDS. Furthermore, it may cause severe sequelae in newborns after prenatal transmission. Intrauterine primary infections are second only to Down’s syndrome as a known cause of mental retardation. Less severe complications are the result of secondary infections (for a review, see reference 6). Diagnosis of HCMV infection can be obtained by direct demonstration of the virus or virus components in pathological materials or indirectly through serology (10). Diagnosis of primary HCMV infection is accomplished exclusively by serological methods, i.e., demonstration of the appearance of antibodies to HCMV in a previously seronegative subject. In the absence of seroconversion, diagnosis can be made by testing serum samples for the presence of HCMV-specific immunoglobulin M (IgM). HCMV-specific IgM is a sensitive and specific indicator of primary HCMV infection in immunocompetent subjects (17). HCMV-specific IgM can also be produced during active viral reactivation in transplant recipients (1, 3, 9). However, its detection has been hampered by a variety of technical problems causing interassay variability. IgM to HCMV can be detected by a variety of different procedures, enzyme immunoassay (EIA) being the most widely used. Many different HCMV IgM EIAs are commercially available for detection of serum IgM. Unfortunately, poor agreement has been found among the results obtained with different kits when the same population of sera was tested (14, 19). Western blots (immunoblots) (WBs) which contain proteins from purified viral particles (viral WBs) have been shown to be a reliable and sensitive method of detecting HCMV-specific IgM (1, 5, 13). However, virological diagnosis, i.e., the direct demonstration

* Corresponding author. Mailing address: Department of Clinical and Experimental Medicine, Section of Microbiology, S. Orsola General Hospital, Via Massarenti 9, 40138 Bologna, Italy. Phone: 39.51.341652. Fax: 39.51.341632. Electronic mail address: TEL1811 @IPERBOLE.BOLOGNA.IT. 597

598

CLIN. DIAGN. LAB. IMMUNOL.

NOTES

TABLE 1. Comparison of IgM detection by viral WB and conv-EIA No. of serum samples with the indicated viral WB result

convEIA result

Positive

Negative

Total

Positive Negative Total

181 171 352

5 360 365

186 531 717

body reaction was observed or if an antibody reaction was observed only with proteins with molecular masses other than 150, 65, 55, or 38 kDa. The virological methods used to identify acute phases of HCMV infection were virus isolation in culture, detection of antigenemia, and detection of viral genome by PCR. The shell vial procedure (4) was used for HCMV isolation from urine and saliva specimens of immunocompetent subjects. The inoculated cells were fixed 24 to 48 h after inoculation and stained in an indirect immunofluorescence assay using a monoclonal antibody reacting with the HCMV IE1/IE2 gene product (E13 from Argene, Varilhes, France). The presence of HCMV pp65 (ppUL83) in PMNL of immunosuppressed patients (antigenemia) was determined as originally described by van der Biji et al. (23) and modified by Revello et al. (20), using an HCMV pp65-specific pool of two monoclonal antibodies (1C3 and AYM-1 from Argene) in indirect immunofluorescence tests. The presence of HCMV genome in PMNL of immunocompromised subjects was detected by PCR. Aliquots of 5 3 105 PMNL were resuspended in 100 ml of PCR buffer (50 mM KCl, 10 mM Tris-HCl [pH 8.3], 2 mM MgCl2, 0.01% gelatin) containing nonionic detergents and proteinase K. Samples were incubated at 608C for 1 h and then at 958C for 20 min to inactivate the proteinase. Subsequently, 30 ml of each sample was added to 20 ml of reaction buffer containing 50 mM KCl, 10 mM Tris-HCl (pH 8.3), 2 mM MgCl2, the four deoxynucleotide triphosphates (0.2 mM each), 50 pmol of each primer, and 1.25 U of native Taq DNA polymerase (Perkin-Elmer Cetus, Norwalk, Conn.). The HCMV-specific primers from the fourth exon of the HCMV immediate-early gene corresponding to nucleotides 1767 to 1786 and nucleotides 1894 to 1913 were used (8). With these primers, a fragment of 147 bp was amplified. A third oligonucleotide consisting of nucleotides 1807 to 1847, which was complementary to the antisense DNA strand in the region between the binding sites of the other oligonucleotides, was used for hybridization (15). The sensitivity (one viral genome) and specificity of the PCR were determined as described in a previous work (15). The results obtained by testing the sera for the presence of HCMV-specific IgM by viral WB were compared with those obtained by conv-EIA. As shown in Table 1, 541 cases of a total of 717 gave concordant results (75.4% agreement between the two tests). Therefore, approximately 25% of the serum samples gave discordant results when tested by the two different serological procedures. This finding is not surprising in view of the other data present in the literature indicating a low level of agreement between IgM detection with different kinds or even the same kind of serological procedure (9, 14, 19). When we analyzed the detailed reactivity to individual HCMV proteins, we observed that approximately 50% of the viral-WB-positive and conv-EIA-negative serum samples showed reactivity either to ppUL32 (pp150) alone (25%) or to pp150 and ppUL80a (pp38) (33%). The remaining reactivities were widely distributed among different combinations of proteins (data not shown).

TABLE 2. Comparison of IgM detection by conv-EIA and the presence of the virus in pathological materials by virological methods Virological diagnosis


No. of serum samples with the indicated conv-EIA result Positive

Negative

Total

186 0 186

233 298 531

419 298 717

The results obtained with the two serological procedures were then compared with results obtained by virological detection, which is the reference procedure for determining an active infection. As shown in Table 2, 484 of 717 samples gave concordant results (agreement, 67.5%) by conv-EIA and virological tests. The poor agreement between IgM detection by conv-EIA and virological results was observed for both immunocompetent (76.7%) and immunocompromised subjects (63.7%). The difference observed between these groups of subjects is not statistically significant (x2 5 3.44). The agreement observed between IgM detected by viral WB and the results obtained by virological detection of HCMV was greater than the agreement of IgM detected by conv-EIA and virological results. In fact, as shown in Table 3, 636 of 717 cases gave concordant results (88.7% agreement). This result is in agreement with results recently obtained by Kraat et al. (9) for renal transplant recipients. Those researchers found an 83% agreement between IgM detected by viral WB and isolation of virus from blood. In contrast to what we observed with convEIA, when the viral-WB data were compared with virological detection of HCMV, better agreement was found for immunocompromised subjects (91.3%) and immunocompetent individuals (86.2%). However, the difference observed between these groups of subjects was not statistically significant (x2 5 0.84). The reason the agreement between viral-WB and virological data is better than that between conv-EIA and virology is the greater sensitivity of the viral WB, since the specificities of the viral WB and conv-EIA are comparable (100%). This is in agreement with the results obtained in a previous work (14) in which the sensitivity and specificity of the commercial kit used in the present work were evaluated and found to be 49.3 and 97.6%, respectively. When we analyzed the detailed reactivity to HCMV proteins in relationship with a positive virological diagnosis (Table 4), we observed that pp150 is the protein which is recognized most frequently by IgM (90.5% of IgM-positive samples), followed by pp38 (65.6%) and ppUL83 (pp65) (29.2%). These results are in agreement with many data present in the literature (7, 11, 13, 18, 21). Furthermore, 62 serum specimens (14.5%) from patients with a positive virological diagnosis were IgM

TABLE 3. Comparison of IgM detection by viral WB and the presence of the virus in pathological materials by virological methods Virological diagnosis


No. of serum samples with the indicated viral WB result Positive

Negative

Total

349 3 352

78 287 365

427 290 717

VOL. 3, 1996

NOTES

599

62

0

92

0

2

0

3

0

15

0

2

0

16

0

25

0

3

0

8

0

18

0

26

0

1

0

10

0

1

0

4

0

15

0

1

0

1

0

3

0

8

1

13

0

2

0

11

0

5

0

1

8

0

0

1

TABLE 4. Detailed analysis of the viral-WB results in relationship to the presence of the virus in pathological materials as detected by virological procedures

70

2

No. of serum samples with the indicated viral protein(s)a recognized by IgM in viral WB Virological 150, 82, 150, 82, 150, 82, 150, 82, 150, 82, diagnosis 82, 150, 150, 150, 150, 150, 55, 150, 150, 150, 150, 65, 150, 150, 150, 82, 150, 82, 150, 150, 82, 55, 65, 65, (n) None 150 55, 38, 65, 38, 65, 55, 65, 55, 65, 55, 38 65 82b 65 38 38, 28 55 55, 38 38, 28 65 65, 38 65, 55 55, 38 82 82, 38 38, 28 55, 38 82, 65 65, 38 38 38 55 28 28 28 38 38, 28

Positive (419)

287

Proteins are listed according to their molecular weights (in thousands); e.g., 150, pp150; 38, pp38; etc. Samples which reacted with p82 protein exclusively were counted as negative for IgM by viral WB in Table 3.

Negative (298)

a

b

reactive exclusively to pp150, 13 serum samples (3%) were IgM reactive exclusively to pp38, and 11 (2.6%) samples were IgM reactive exclusively to pp65 on the viral WB. These data support the fact that pp150, pp38, and pp65 are important antigens for IgM detection. Table 4 also shows that 8 of 11 serum samples that were judged IgM positive by viral WB and negative by virological detection reacted exclusively with a protein with a molecular mass of 82 kDa (pp82). The reactivity to pp82 alone was observed for seven blood donors and one renal transplant recipient and never observed alone among the serum samples from patients with an ongoing HCMV infection. Therefore, the serum IgM reactivity to pp82 alone should not be considered a positive reaction. Furthermore, two serum specimens showed reactivity to pp150 alone and one serum sample reacted only with pp38. As pp150 is the strongest HCMV immunogen and pp38 preferentially induces an IgM response without inducing any significant switching to IgG, it is likely that IgM to these proteins can persist for a long time after the end of the acute phase of viral replication. However, we cannot exclude the possibility that the reactivity to pp150 is not due to ppUL32 but, rather, to the other structural protein with a molecular mass of 150 kDa (pUL86) which is conserved among herpesviruses. In addition, a nonspecific reaction to pp38 could be due to some cross-reactivity of this protein with a cell membrane protein of 60 kDa (12). Therefore, some serum samples that are IgM reactive only to 150- or 38-kDa proteins on the viral WB may not be due to HCMV-specific IgM. The reactivity to pp150 and to pp38 alone should be considered with caution. In conclusion, the viral WB is a serological procedure more sensitive than conv-EIA (Enzygnost anti-HCMV/IgM kit) for anti-HCMV IgM detection, and the results obtained correlate much better than those obtained by conv-EIA with an active HCMV infection as documented virologically for both immunocompetent and immunosuppressed subjects. Data present in the literature indicating a correlation between serology and virology lower than that obtained in this work (5, 16) can be explained by the different serological procedures used to detect HCMV-specific IgM as well as the less sensitive virological procedures used to detect the virus. REFERENCES 1. Basson, J., J. C. Tardy, and M. Aymard. 1989. Pattern of anti-cytomegalovirus IgM antibodies determined by immunoblotting. A study of kidney graft recipients developing a primary or recurrent CMV infection. Arch. Virol. 108:259–270. 2. Boland, G. J., G. C. deGast, R. J. Hene´, G. Jambroes, R. Donckerwolcke, T. H. The, and G. C. Mudde. 1990. Early detection of active cytomegalovirus (CMV) infection after heart and kidney transplantation by testing for immediate early antigenemia and influence of cellular immunity on the occurrence of CMV infection. J. Clin. Microbiol. 28:2069–2075. 3. Dolan, J., J. D. Briggs, and G. B. Clements. 1989. Antibodies to cytomegalovirus in renal allograft recipients: correlation with isolation of virus. J. Clin. Pathol. 42:1070–1077. 4. Gleaves, C. A., T. F. Smith, E. A. Shuster, and G. R. Pearson. 1984. Rapid detection of cytomegalovirus in MRC5 cells inoculated with urine specimens by use of low-speed centrifugation and monoclonal antibody to an early antigen. J. Clin. Microbiol. 19:917–919. 5. Gold, D., R. Ashley, H. H. Handsfield, M. Verdon, L. Leach, J. Mills, L. Drew, and L. Corey. 1988. Immunoblot analysis of the humoral immune response in primary cytomegalovirus infection. J. Infect. Dis. 157:319–325. 6. Ho, M. 1991. Cytomegalovirus. Biology and infection, 2nd ed. Plenum Medical Press, New York. 7. Jahn, G., B. C. Scholl, B. Traupe, and B. Fleckenstein. 1987. The two major phosphoproteins (pp65 and pp150) of human cytomegalovirus and their antigenic properties. J. Gen. Virol. 68:1327–1337. 8. Jiwa, N. M., G. W. Gemert, A. K. Raap, F. M. Van der Riijjke, A. Mulder, P. F. Lens, M. M. M. Salimans, F. E. Zwaan, W. Van Dorp, and M. Van der Ploeg. 1989. Rapid detection of human cytomegalovirus DNA in peripheral blood leukocytes of viremic transplant recipients by the polymerase chain reaction. Transplantation 48:72–76. 9. Kraat, Y. J., F. S. Stals, M. H. L. Christiaans, T. Lazzarotto, M. P. Landini,

600

10. 11.

12.

13.

14.

15.

16.

NOTES and C. A. Bruggeman. 1996. IgM antibody detection for 38 (ppUL80a) and 150 (ppUL32) kDa proteins by immunoblotting: the earliest parameter for acute cytomegalovirus infection in renal transplant recipients. J. Med. Virol. 48:289–294. Landini, M. P. 1993. New approaches and perspectives in cytomegalovirus diagnosis. Prog. Med. Virol. 4:157–177. Landini, M. P., M. X. Guan, G. Jahn, W. Lindenmeier, M. Mach, A. Ripalti, A. Necker, T. Lazzarotto, and B. Plachter. 1990. Large-scale screening of human sera with cytomegalovirus recombinant antigens. J. Clin. Microbiol. 28:1375–1379. Landini, M. P., T. Lazzarotto, E. Percivalle, and G. Gerna. 1991. Evidence that human cytomegalovirus assembly protein shares antigenic sites with an uninfected cell membrane protein. J. Gen. Virol. 72:3009–3016. Landini, M. P., E. Rossier, and H. Schmitz. 1988. Antibodies to human cytomegalovirus structural polypeptides during primary infection. J. Virol. Methods 22:309–317. Lazzarotto, T., B. Dalla Casa, B. Campisi, and M. P. Landini. 1992. Enzymelinked immunoadsorbent assay for the detection of cytomegalovirus-IgM: comparison between eight commercial kits, immunofluorescence and immunoblotting. J. Clin. Lab. Anal. 6:216–218. Lazzarotto, T., G. Furlini, M. C. Re, E. Ramazzotti, B. Campisi, and M. P. Landini. 1994. Human cytomegalovirus replication correlates with differentiation in a hematopoietic progenitor cell line and can be modulated by HIV. Arch. Virol. 135:13–28. Marsano, L., R. P. Perrillo, M. W. Flye, D. W. Hanto, E. D. Spitzer, J. R. Thomas, P. R. Murray, D. W. Windus, E. M. Brunt, and G. A. Storch. 1990. Comparison of culture and serology for the diagnosis of cytomegalovirus

CLIN. DIAGN. LAB. IMMUNOL. infection in kidney and liver transplant recipients. J. Infect. Dis. 161:454–461. 17. Nielsen, S. L., I. Sorensen, and H. K. Andersen. 1988. Kinetics of specific immunoglobulins M, E, A, and G in congenital, primary, and secondary cytomegalovirus infection studied by antibody-capture enzyme-linked immunosorbent assay. J. Clin. Microbiol. 26:654–661. 18. Plachter, B., L. Wieczorek, B. C. Scholl, R. Ziegelmaier, and G. Jahn. 1992. Detection of cytomegalovirus antibodies by an enzyme-linked immunosorbent assay using recombinant polypeptides of the large phosphorylated tegument protein pp150. J. Clin. Microbiol. 30:201–206. 19. Re, M. C., and M. P. Landini. 1989. IgM to human cytomegalovirus: comparison of two enzyme immunoassays and IgM reactivity to viral polypeptides detected by immunoblotting. J. Clin. Lab. Anal. 3:169–173. 20. Revello, M. G., M. Zavattoni, E. Percivalle, P. Grossi, and G. Gerna. 1989. Correlation between immunofluorescent detection of human cytomegalovirus immediate early antigens in polymorphonuclear leukocytes and viremia. J. Infect. Dis. 160:159–160. 21. Scholl, B. C., B. Von Hintzestein, B. Borisch, B. Traupe, M. Broker, and G. Jahn. 1988. Procaryotic expression of immunogenic polypeptides of the large phosphoprotein (pp150) of human cytomegalovirus. J. Gen. Virol. 69:1195– 1204. 22. The`, T. H., A. P. van der Berg, W. J. van Son, A. M. Tegzess, M. J. H. Sloof, I. J. Klompmaker, E. B. Haagsma, W. van der Biji, and M. van der Giessen. 1991. Recent advances in the early and reliable immunodiagnosis of cytomegalovirus infection in immunocompromised hosts, p. 209–220. In M. P. Landini (ed.), Progress in cytomegalovirus research. Elsevier, Amsterdam. 23. van der Biji, W., R. Torensma, W. J. Son, J. Schirm, A. M. Tegzess, and T. H. The. 1988. Rapid immunodiagnosis of active cytomegalovirus infection by monoclonal antibody staining of blood leukocytes. J. Med. Virol. 25:179–188.