To better characterize cytomegalovirus (CMV) isolates in a pediatric population, we performed glycoprotein B genotyping of isolates from CMV culture-positive patients during a 3-year period at Children’s Hospital Medical Center, Cincinnati, Ohio. The most common sources for CMV isolation were respiratory and urine cultures. Among 61 isolates glycoprotein B genotype distribution was: genotype 1, 32%; genotype 2, 24%, genotype 3, 32%; and genotype 4, 10%. Among viremic isolates there was a trend favoring type 1 genotype (P = 0.057).
Although cytomegalovirus (CMV) infection is generally asymptomatic in immunocompetent patients, severe disease may occur in immunocompromised hosts, including organ transplant recipients, HIV-infected individuals and infants. 1 Because of the critical role children play in the transmission of CMV infection, a better understanding of the molecular epidemiology of this infection in pediatric patients may be important for devising control strategies. Limited molecular analyses of clinical CMV isolates have revealed considerable genetic heterogeneity among strains. The glycoprotein B (gB) coding sequence exhibits significant genetic polymorphism, providing the basis for a classification scheme referred to as “gB genotyping.” A number of studies have investigated the correlation of gB genotypes with CMV disease manifestations in adults. 2–5 This study examined correlation of CMV gB genotype with disease syndrome in pediatric patients.
Materials and methods.
Children’ Hospital Medical Center, Cincinnati, is a 335-bed regional hospital. All primary isolates of CMV are saved in frozen storage. Sequential stored isolates (n = 84) from a 4-year period (1995 to 1998) were subcultured onto human foreskin fibroblasts. After identification of characteristic cytopathic effect, monolayers were harvested and DNA was purified using the Qiamp DNA purification system (Qiagen) according to the manufacturer’s specifications. The laboratory CMV isolate, Towne (the gift of A. P. Geballe), was cultivated as a control. PCR was performed as described elsewhere 2 with a denaturation cycle at 96°C for 6 min, followed by 35 cycles of denaturation for at 96°C for 1.5 min, annealing at 55°C for 2 min and extension at 72°C. PCR amplification products were divided into separate aliquots and subjected to endonuclease digestion with 10 units of Rsa I and Hin f 1, respectively, overnight at 37°C. Restriction digests were subjected to electrophoresis in 3.5% agarose gels in 1× tris(hydroxymethyl)aminomethane-borate-EDTA buffer. Agarose gels were stained with ethidium bromide, visualized and photographed with the use of an Alpha-Innotech Digital Image Analyzer. Statistical comparisons were conducted with Fisher’s exact test.
In the period from 1995 to 1998, 84 viral isolates from 77 patients could be identified. From these, 63 unique isolates could be successfully recultured from frozen storage. The majority of isolates (44%) were obtained from salivary/respiratory tract cultures, whereas urine culture accounted for 30% of the cultured isolates. Blood culture was the next most common source for CMV in our population, with 15 viremic isolates being identified (24% of all subcultured isolates). Medical records were available for review for 58 of these patients. A high percentage of viral isolates was from immunocompromised patients, with 21 patients (33%) identified as solid organ transplant patients, bone marrow/stem cell transplant patients or oncology patients. The second most common category was congenital CMV infection (12 infants). In 30 instances (48%) there were no discernable risk factors for CMV infection, or the medical record was not available for review. For these patients it was determined that CMV infection was an incidental finding not related to hospitalization. This group included children with fever of unknown origin, cystic fibrosis patients or children with acute respiratory infections (bronchiolitis, community-acquired pneumonia). No CMV isolates were identified in children with HIV infection.
Of 63 CMV isolates 61 could be successfully gB genotyped (97%). Using restriction fragment length polymorphism genotyping, 32% of isolates were classified as genotype 1, 24% as genotype 2, 32% as genotype 3 and 10% as genotype 4. The distribution of isolates was examined as a function the patient disease category (Table 1). Based on chart review disease categories were separated into oncology/transplant patients, infants with congenital CMV infection or children with no evidence of active CMV disease. When the analysis was limited to transplant patients (solid organ, bone marrow or stem cell transplant patients) or oncology patients, the gB genotype distribution was similar to that for the overall group, with equal numbers of genotypes 1 and 3 (6 of 21 for each genotype) observed. In contrast among infants with congenital infection, there was a trend favoring genotype 1 [6 of 12 (50%) compared with 20 of 63 (32%) for all CMV isolates (Table 1)]. However, this difference was not statistically significant (P = 0.17 vs. noncongenital infection isolates). To next analyze whether certain gB genotypes correlated with a propensity to induce viremia, we compared blood isolates with non-blood isolates. Among viremic isolates, genotype 1 isolates also predominated; 8 of 15 blood isolates (53%) were type 1, compared with 12 of 48 type 1 genotypes (25%) for non-blood isolates (Table 2;P = 0.057, Fisher’s exact test).
In this study we examined the gB genotype distribution of wild-type CMV isolates obtained in a pediatric population, using restriction fragment length polymorphism analysis and found it to be very similar to that reported among adult patients in previous series. Similar to previously reported series in adults, 2–5 gB genotype 1 was the predominant strain, representing 34% of all isolates. Genotype 2 accounted for 25% of all isolates, and genotype 3 represented 32%. As has been observed in some previous reports, genotype 4 isolates were relatively uncommon, representing only 9% of our isolates. Our results are very similar to the distribution originally reported by Chou and Dennison 2 (38% type 1, 28% type 2, 28% type 3 and 8% type 4) but differ somewhat from the data reported by Meyer-Konig et al., 3 who identified fewer type 3 isolates in their series (36% type 1, 32% type 2, 16% type 3 and 16% type 4). The differences described in various reported series may depend on the patient population examined or the source of the viral isolate.
It remains unclear to what extent the variability in CMV genes, including gB, correlates with severity of CMV manifestations. In subsets of patients with HIV infection, including patients with AIDS-associated CMV retinitis, genotype 2 has been identified as the predominant strain, 6, 7 whereas in a recently reported series in liver transplant patients, genotype 3 predominated. 8 In a study of aplastic anemia, gB genotype 3 was more likely to be identified in bone marrow aspirates from these patients, suggesting a causative link between this genotype and the development of bone marrow aplasia. 9 Thus variation in gB genotype may predict the propensity for opportunistic CMV disease. There is limited information about gB genotypes and congenital CMV disease. In recent reports from Europe, gB genotypes 1 and 3 have been suggested to be associated with adverse outcomes in the setting of congenital infection. 10, 11 Data from Texas and Iowa correlating gB genotype in congenitally infected infants with neurodevelopmental outcome suggested that genotype 3 was more commonly associated with asymptomatic infection. 12 Our data suggest an association between gB genotype 1 and CMV viremia, although this trend did not achieve statistical significance. Because infection in utero is thought to reflect hematogenous transmission of virus transplacentally to the fetus, it is conceivable that strains with increased propensity for viremia could predispose to congenital transmission. Further molecular epidemiologic studies are necessary to assess whether variation in gB genotype or other virulence genes 13, 14 correlates with the outcome of CMV infection.
The technical assistance of Nancy Jensen is acknowledged. This work was supported by National Institutes of Health Grants AI-65289, HD38416-01 and AI-45252 and by a grant from the Children’s Hospital Research Foundation (to SAX).
1. Schleiss MR. Cytomegalovirus
. In: Rudolph CD, Rudolph AM, Hostetter MK, Lister G, Siegel NJ, eds. Rudolph’s pediatrics. 21st ed. New York: McGraw-Hill, 2003: 1031–5.
2. Chou SW, Dennison KM. Analysis of interstrain variation in cytomegalovirus glycoprotein B
sequences encoding neutralization related epitopes. J Infect Dis 1991; 163: 1229–34.
3. Meyer-Konig U, Vogelberg C, Bongarts A, et al. Glycoprotein B
genotype correlates with cell tropism in vivo of human cytomegalovirus
infection. J Med Virol 1998; 55: 75–81.
4. Fries BC, Chou S, Boeckh M, Torok-Storb B. Frequency distribution of cytomegalovirus
envelope glycoprotein genotypes in bone marrow transplant recipients. J Infect Dis 1994; 169: 769–74.
5. Rasmussen L, Hong C, Zipeto, et al. Cytomegalovirus
gB genotype distribution differs in human immunodeficiency virus-infected patients and immunocompromised allograft recipients. J Infect Dis 1997; 175: 179–84.
6. Fidouh-Houhou N, Duval X, Bissuel F, et al. Salivary cytomegalovirus
(CMV) shedding, glycoprotein B
genotype distribution, and CMV disease in HIV-seropositive patients. Clin Infect Dis 2001; 33: 1406–11.
7. Drew WL, Chou S, Miner RC, et al. Cytomegalovirus glycoprotein B
groups in human immunodeficiency virus-infected patients with incident retinitis. J Infect Dis 2002; 186: 114–7.
8. Sarcinella L, Mazzulli T, Willey B, Humar A. Cytomegalovirus glycoprotein B
genotype does not correlate with outcomes in liver transplant patients. J Clin Virol 2002; 24: 99–105.
9. Torok-Stob B, Bolles L, Iwata M, et al. Increased prevalence of CMV gB3 in marrow of patients with aplastic anemia. Blood 2001; 98: 891–2.
10. Barbi M, Binda S, Caroppo S, et al. CMV gB genotypes and outcome of vertical transmission: study on dried blood spots of congenitally infected babies. J Clin Virol 2001; 21: 75–9.
11. Lukacsi A, Tarodi B, Endreffy E, et al. Human cytomegalovirus
gB genotype 1 is dominant in congenital infections in South Hungary. J Med Virol 2001; 65: 537–42.
12. Bale JF, Murph JR, Demmler GJ, et al. Intrauterine cytomegalovirus
infection and glycoprotein B
genotypes. J Infect Dis 2000; 182: 933–6.
13. Bale JF, Petheram SJ, Robertson M, et al. Human cytomegalovirus
a sequence and UL144 variability in strains from infected children. J Med Virol 2001; 65: 90–6.
14. Arav-Boger R, Willoughby RE, Pass RF, et al. Polymorphisms of the cytomegalovirus
(CMV)-encoded tumor necrosis factor-alpha and beta-chemokine receptors in congenital CMV disease. J Infect Dis 2002; 186: 1057–64.