1629 Presence of Human Immunodeficiency Virus (HIV) Type 1 Subtype A Infection in a New York Community with High HIV Prevalence: A Sentinel Site for Monitoring HIV Genetic Diversity in North America Kathleen L. Irwin, Chou-Pong Pau, Davis Lupo, Danuta Pienazek, Chi-Cheng Luo, Noemi Olivo, Mark Rayfield, Dale J. Hu, J. Todd Weber, Richard A. Respess, Robert Janssen, Patrick Minor, Jerome Ernst, and the Centers for Disease Control and Prevention Bronx Lebanon HIV Serosurvey Team Centers for Disease Control and Prevention, and Orkand Corporation, Atlanta; Bronx Lebanon Hospital Center, Bronx, New York To determine whether US residents are infected with subtypes of human immunodeficiency virus (HIV) type 1 other than subtype B (), the predominant North American subtype with a unique GPGR genetic sequence in the V3 loop, viruses from 22 HIV-infected adults were serotyped and subtyped. Twenty patients had subtype B (), of whom 15 had serotype B (), 3 had serotype A/C, 1 had serotype B (Thai), and 1 had a nontypeable serotype. Two had subtype A, both serotype A/C. Both subtype A infected patients, only 1 of whom had been outside the United States, reported sex with persons traveling abroad, suggesting possible acquisition in the United States. Because US residents are infected with non subtype B () strains, US surveillance for HIV-1 diversity is needed to elucidate subtype-specific transmission patterns and pathogenesis and to guide evaluation and development of HIV diagnostic tests and vaccines. To date, the vast majority of human immunodeficiency virus Materials and Methods type 1 (HIV-1) isolates reported from North America and Europe contain a unique GPGR genetic sequence in the V3 loop During 1992 1994, we studied inpatients and outpatients tip and are classified as group M, subtype B, referred to in this treated at Bronx-Lebanon Hospital Center who met these eligibility report as subtype B () [1]. However, rare cases of criteria: aged 18 44 years; admitted to medicine, surgery, gynecolinfection with HIV-1 subtypes A, D, and E and group O have ogy, or family practice wards or triaged to the nonurgent emer- been reported in immigrants [2, 3] and in US servicemen who gency department track; not known to be HIV-infected before enrollment; not admitted for AIDS-related conditions; not offered were infected while living overseas [4]. Little is known about HIV testing by nonstudy staff during this admission; physically and the prevalence of non subtype B () subtypes in wellmentally able to complete interviews; and able to speak English or characterized US populations and whether non subtype B Spanish. Eligible patients (n Å 1749) were invited to participate () strains are being transmitted among US residents in voluntary HIV counseling and testing and pre- and posttest who have not traveled abroad. To examine the diversity of interviews about HIV risk behaviors and travel history. Nearly HIV-1 subtypes among US residents, we examined patients half (n Å 828, 47.3%) agreed to testing and interviews. Information with newly identified HIV-1 infection who resided in the South on characteristics of patients who declined interview was not avail- Bronx, New York, a community having one of the highest able from other sources. All HIV-infected participants were newly prevalences of HIV infection and highest proportions of immi- diagnosed. grants and visitors from the Caribbean, Latin America, and Patient sera were tested for HIV antibodies (HIV-1/HIV-2 im- Africa in the United States [5]. munoassay; Genetic Systems, Seattle) and confirmed by blot (Cambridge Biotech, Worcester, MA). The 43 blotpositive specimens were serotyped using investigational EIAs including 10 peptides representing group M subtype V3 loop immunogenic domains of subtypes A, B (), B (Thai), B (Brazilian), C, D, E, and F and group O subtype (Ant-70 and Submitted 27 February 1997; revised 29 May 1997. Presented in part: XI International Conference on AIDS, Vancouver, British MVP5180) [6]. Specimens reactive to V3-A or V3-C peptide Columbia, 10 July 1996 (abstract We.C.345). immunoassays were classified as serotype A/C because these im- Financial support: CDC Cooperative Agreement U64/CCU206797. Informed consent was obtained from all patients, and human experimentation munoassays are not highly specific for subtypes A and C [7]. guidelines of the US Department of Health and Human Services and Bronx- Specimens were also serotyped using five peptides derived from Lebanon Hospital Center were followed in the conduct of this research. HIV-1 gp41 transmembrane proteins (groups M and O) and one Reprints or correspondence: Dr. Kathleen Irwin, Division of HIV/AIDS peptide derived from HIV-2 gp36 protein [6]. Prevention, Centers for Disease Control and Prevention, 1600 Clifton Rd., Viruses were genetically sequenced from proviral DNA obtained Mail Stop E-45, Atlanta, GA 30333. from peripheral blood mononuclear cells [8] or viral RNA ex- The Journal of Infectious Diseases 1997;176:1629 33 1997 by The University of Chicago. All rights reserved. tracted from serum using QIAamp blood kits (Qiagen, Chatsworth, 0022 1899/97/7606 0032$02.00 CA). Samples were amplified using polymerase chain reaction
1630 Concise Communications JID 1997;176 (December) (PCR) [9] according to manufacturer s specifications (Perkin Elmer, Foster City, CA). Several primers (Abbott Laboratories, Abbott Park, IL) were used to generate a 525-nucleotide fragment of the C2-V3 envelope region: for primary PCR, JH44F (5 -ACA- GTRCARTGYACACATGG, nt 6983 7002) and JH35mR (5 - CACTTCTCCAATTGTCCITCA, nt 7668 7688); for nested PCR, JH33F (5 -CTGTTIAATGGCAGICTAGC, nt 7031 7050) and JH48R (5 -RATGGGAGGRGYATACAT, nt 7538 7555). When required, viral RNA was reverse-transcribed using primer JH35mR. Nucleotide positions were based on HIV MN complete genome Genbank accession number M17449. All PCR products were purified (QIAquick purification kits; Qiagen) and were subjected to automated sequence analysis using dye terminator cyclesequencing reagents and protocols (ABI PRISM; Perkin Elmer). Phylogenetic analysis was performed using neighbor-joining methods with bootstrap resampling [10]. denied foreign birthplace or travel but reported having had sex with partners who had spent time in the Caribbean, Latin America, or Africa. Patient BX912 was born in Africa but declined to provide information on the age at which he moved to the United States or the country in which sexual activities with persons who had spent time in the Caribbean, Latin America, or Africa had taken place. Both patients infected with subtype A (BX866 and BX912) reported having recently practiced unprotected heterosexual sex. Both denied having received transfusions before 1985. The demographic and HIV risk characteristics of the 2 patients with subtype A infection and the patients with confirmed subtype B () infection were not statistically different at P õ.05 using Fisher s exact test (data not shown). Discussion Results Of the 43 newly identified HIV-infected hospital patients, 2 Of the 828 patients tested for HIV-1, 43 (5.2%) were seropositive; (4.7%) were infected with HIV-1 subtype A. We may have failed none of the 43 had serotype or genotype consistent to detect other non subtype B () strains in this commu- with HIV-1 group 0 or HIV-2 infection. Thirty-five (81.4%) nity because we enrolled only English and Spanish speakers and of 43 patients had serotype B (). C2-V3 envelope re- because patients, including Africans, who spoke only French or gions were sequenced from 15 of these 35 patients, of whom African languages, were ineligible. However, õ20 patients ap- 6 were foreign-born and 9 were US-born patients who denied proached for the study were ineligible for this reason. travel outside the continental United States and sex with persons To our knowledge, this is the first report of subtype A infection who had spent time in the Caribbean, Latin America, or in an American-born US resident who had not traveled outside the Africa. All 15 had subtype B () (figure 1). Eight of the United States. The epidemiologic characteristics of the 2 patients 43 (18.6%) patients had non B () serotypes: 6 infected with subtype A indicate that they may have acquired (BX866, BX912, BX269, BX536, BX658, and BX134) had their infections while residing in the United States, although non serotype A/C, 1 (BX130) had serotype B (Thai), and 1 (BX500) locally acquired infection cannot be ruled out for the patient born was classified as nontypeable because the patient s serum reacted in Africa. Three patients with serotype A/C infected with subtype strongly to the gp41-lai peptide characteristic of HIV- B () strains had V3 loop sequences that were atypical of 1 group M but lacked reactivity to all V3 peptides (table 1). the subtype B () strains most commonly recovered in the C2-V3 envelope regions were sequenced from 7 of these 8 United States and thus reacted more strongly to the serotype A patients, for whom we had adequate specimen volume. Two immunoassays than to the serotype B () immunoassays. patients classified as serotype A/C were infected with viruses As others have reported [11 14], this may be due to infection of subtype A sequences (BX866 and BX912), and 5 patients with 2 virus strains (A and B []), a recombinant strain, were infected with HIV subtype B () (figure 1). The or a subtype B () strain with a V3 loop sequence that immunogenic domain of the V3 region of the virus infecting differed substantially from other US strains that have been se- patient BX130 (GPGQAWY) was more similar to B (Thai) quenced. However, additional sequencing would be needed to strains (GPGQAWY) than to other sequenced subtype B strains confirm any of these explanations. Patient BX500, who had sub- (GPGRAFY) and thus was classified as serotype B (Thai). type B () virus but was nonreactive to all 10 V3 loop However, more extensive phylogenetic analysis of the C2-V3 peptides, may have lacked measurable antibody to the V3 loop, envelope region sequence revealed only a weak association a finding reported in seroconverters infected with subtype B (characterized by very low, unstable bootstrap values) with () viruses [15, 16]. This and other studies [16] demonstrate sequences from 16 B (Thai) strains (data not shown). This that serotype based on these V3 loop and gp41 peptides and analysis, therefore, suggested that the patient s virus was most subtyping by sequencing analysis does not always agree. How- consistent with subtype B () (figure 1). Patient BX500, ever, serotyping was useful to distinguish patients with subtype who lacked reactivity to all V3 loop peptides, had an unusual B () viruses with typical sequences from patients with GPGKV motif within the V3 loop sequence instead of the non B () viruses or B () viruses with unusual GPGRA motif typical of subtype B () strains. sequences. The 7 patients with non B () serotypes whose vi- Our report and others [2 4] of US residents infected with ruses underwent genotyping had diverse characteristics (table non subtype B () viruses indicate that multiple intro- 1). Of note, patient BX866, who was infected with subtype A, ductions of HIV-1 have occurred in North America and will
JID 1997;176 (December) Concise Communications 1631 Figure 1. Phylogenetic classification of HIV-1 strains of 22 patients from Bronx, New York. Nos. to left side of branch nodes indicate probabilities based on bootstrap resampling evaluations of 100 data sets. Nos. to right of branch nodes indicate patient identification numbers; Bronx patients are denoted with BX prefix. Patients infected with subtype A viruses are marked with arrowheads. Classification of distant HIV-1 subtypes A H are delineated by uppercase letters on right. Sequence information has been deposited in GenBank (accession nos. U90181 U90202).
1632 Concise Communications JID 1997;176 (December) Table 1. Selected characteristics of patients with newly identified HIV-1 infection with non subtype B () serotype, Bronx, New York, 1992 1994. High-risk Patient Years in Travel outside Sex partner with heterosexual Male-to- Ever no. Serotype Genotype Sex Birthplace Bronx 50 US states foreign travel* behaviors male sex injected BX866 A/C A F New York City 10 No Yes Yes NA No BX912 A/C A M Gambia?? Yes? No No BX269 A/C B F New York City 8 No No Yes NA Yes BX536 A/C B M Continental US 10 No Yes Yes No Yes BX658 A/C B M Puerto Rico 10 Puerto Rico Yes No Yes No BX134 A/C Not done M Mali 2 Burundi, Yes Yes No No Canada BX130 B Thai B M Puerto Rico 10 Puerto Rico No Yes No No BX500 Nontypeable B M New York City?? Yes? Yes No NOTE. M, male; F, female; NA, not applicable, female patient;?, unknown; US, United States. * Defined as partner who was born in or spent time in Caribbean, Latin America, or Africa. Defined as having engaged in commercial sex work or sexual contact with commercial sex worker, an HIV-infected person, drug injector, or transfusion recipient, or for women, bisexual man. We are indebted for the invaluable contributions of study team members not listed as authors: Stephen Bekoe-Tabiri, Franklin Benjamin, Stere Carniciu, Carlos Chavaria, Vicsa Chiriboga, Maura Disla, Auria Fernandez, Camelia Ganea, Gilbert Martinez, Reineisse Parsons, Felipe Rosario, William Rothman, Wanda San- tos, Rita Shulman, and Yvette Vasquez, from Bronx-Lebanon Hospital Center; and Karen Brown, Faye Cowart, Avis Daugharty, Marty Edeline, Ernest Lewis, Brent McRae, Vincent Raimondi, Charles Schable, Gerald Schochetman, Larry Wells, and Roy Win- ter, from the Centers for Disease Control and Prevention. probably continue to occur. The strain diversity in our study population may be due to either recent introduction of new strains from outside the United States or to local transmission of non subtype B () strains established in the community. Because the 2 patients with non subtype B () strains had recently practiced unprotected sex, local transmission of these variants may be ongoing. Surveillance that links genetic, epidemiologic, and clinical characteristics in sentinel North American communities with high HIV incidence such as the Bronx may be useful to elucidate subtype-specific patterns of HIV-1 transmission and pathogenesis and to determine if the genetic diversity of HIV-1 will alter the dynamics of the epidemic in North America. Because serotype and subtype are not always consistent and serotyping is less complex and costly than genetic subtyping with genetic sequencing, serotyping is best used to screen large numbers of specimens for peptide reactivity that is inconsistent with subtype B () strains. Specimens with such atypical reactivity should then be genetically sequenced to confirm subtype. In our study, no demographic or HIV risk characteristic that might be useful for identifying patients who might benefit from genetic subtyping (e.g., the interpretation of some HIV virus load assays may depend on subtype [17]) was signifi- cantly associated with the presence of non subtype B () infection. However, the very small sample size limited the statistical power of this analysis. There is no evidence that the HIV-1 infections characterized by either non B () serotype or non B () subtype identified in this study would have compromised blood safety because all 8 patients tested positive on commercial HIV tests and presumably would have tested positive if they had donated blood. Because all 8 denied having received transfusions before 1985, when US blood banks began screening blood for HIV, transfusion-related acquisition is unlikely. The presence of non subtype B () viruses among US residents may also have important implications for HIV diagnostic test sensitivity, ensuring blood safety, and vaccine development. Most HIV tests used to diagnose infection or screen blood in North America are based primarily on subtype B () strains, and their sensitivity for detecting non subtype B () strains requires further evaluation [1]. If ongoing surveillance indicates a high prevalence of non subtype B () strains in US communities that could benefit greatly from HIV vaccines, the development of polyvalent vaccines for use in North America will be warranted. Acknowledgments References 1. Hu DJ, Dondero TJ, Rayfield MA, et al. The emerging genetic diversity of HIV. JAMA 1996;275:210 6.
JID 1997;176 (December) Concise Communications 1633 2. Gao F, Yue L, Hill SC, et al. HIV-1 sequence subtype D in the United 11. Artenstein AW, VanCott TC, Mascola JR, et al. Dual infection with human States. AIDS Res Hum Retroviruses 1994;10:625 7. immunodeficiency virus type 1 of distinct envelope subtypes in humans. 3. Centers for Disease Control and Prevention. Identification of HIV-1 group O J Infect Dis 1995;171:805 10. infection Los Angeles County, California, 1996. MMWR 1996;45:561 5. 12. Pieniazek D, Janini LM, Ramos A, et al. HIV-1 patients may harbour 4. Brodine SK, Mascola JR, Weiss PJ, et al. Detection of diverse HIV-1 viruses of different phylogenetic subtypes: implications for the evolution genetic subtypes in the USA. Lancet 1995;356:1198. of the HIV/AIDS pandemic. Emerg Infect Dis 1995;1:86 8. 5. Irwin KL, Olivo N, Schable C, et al. Absence of HIV-2 infection in a 13. Robertson DL, Sharp PM, McCutchan FE, Hahn BH. Recombination of U.S. population at high risk. Transfusion 1996;36:731 3. HIV-1. Nature 1995;374:124 6. 6. Pau CP, Hu DJ, Spruill C, et al. Surveillance for human immunodeficiency 14. Xin KQ, Shapshak P, Kawamoto S, et al. Highly divergent env sequences of virus type 1 group O infections in the United States. Transfusion 1996; HIV-1 B subtype with two novel V3 loop motifs detected in an AIDS patient 36:398 400. in Miami, Florida. AIDS Res Hum Retroviruses 1995;11:1139 41. 7. Cheingsong-Popov R, Lister S, Callow D, et al. Serotyping HIV Type 1 15. Holmback K, Kusk P, Hulgaard EF, Bugge TH, Scheibel E, Lindhardt by antibody binding to the V3 loop: relationship to viral genotype. BO. Autologous antibody response against the principal neutralizing AIDS Res Hum Retroviruses 1994;10:1379 86. domain of human immunodeficiency virus type 1 isolated from infected 8. Velleca WM, Palmer DF, Feorino P et al. Isolation, culture, and identifica- humans. J Virol 1993;67:1612 9. tion of human immunodeficiency virus. Atlanta: Centers for Disease 16. Pau CP, Kai M, Holloman-Candal DL, et al. Antigenic variation and serotyp- Control and Prevention, 1991. ing of HIV type 1 from four World Health Organization sponsored HIV 9. Wang AM, Doyle MV, Mark DF. Quantitation of mrna by the polymerase vaccine sites. AIDS Res Hum Retroviruses 1994;10:1369 77. chain reaction. Proc Natl Acad Sci USA 1989;86:9717 21. 17. Dunne AL, Crowe SM. Comparison of branched DNA and reverse tran- 10. Felsenstein J. PHYLIP, Phylogeny Inference Package, version 3.5. Cladistics scriptase polymerase chain reaction for quantifying six different HIV- 1989;5:164 6. 1 subtypes in plasma. AIDS 1997;11:126 27. Cytokine Profiles in Cerebrospinal Fluid of Human Immunodeficiency Virus Infected Patients with Cryptococcal Meningitis: No Leukocytosis despite High Interleukin-8 Levels Wendy Chaka, Robert Heyderman, Innocent Gangaidzo, Valerie Robertson, Peter Mason, University of Zimbabwe Meningitis Group, Jan Verhoef, André Verheul, and Andy I. M. Hoepelman Eijkman-Winkler Institute for Microbiology, Infectious Diseases and Inflammation, and Department of Internal Medicine, Division of Infectious Diseases and AIDS, University Hospital Utrecht, Utrecht, The Netherlands; University of Zimbabwe Medical School, Harare, Zimbabwe Cytokine levels were studied in the cerebrospinal fluid (CSF) of 16 adults with cryptococcal meningitis (CM). Low levels of tumor necrosis factor (TNF)-a and interferon-g, high levels of interleukin (IL)-1b, IL-6, and IL-8, and the presence of IL-10 were documented. There were no significant differences in levels of TNF-a and interferon-g for CM and control patients. Mean CSF levels of IL-1b (139.5 pg/ml), IL-6 (346 pg/ml), IL-8 (1160 pg/ml), and IL-10 (9.27 pg/ml) were significantly (P õ.01) elevated in CM patients compared with levels in control patients. Despite the high CSF levels of IL-8, minimal leukocytosis was seen. Significant correlations between cryptococcal antigen titers and IL-10 levels (r Å.8, P õ.05), protein and cryptococcal antigen titer (r Å.9, P õ.05), and protein and IL-10 levels (r Å.8, P õ.05) were found. Cryptococcosis, the disease caused by the yeast-like fungus fected patients [1 3]. Meningitis is the most frequent complication of cryptococcosis. The clinical presentation of this disease, Cryptococcus neoformans, is a major cause of morbidity and mortality in patients with compromised cell-mediated immuated with nonspecific symptoms, such as headache and fever which is often fatal, may be insidious and is commonly associ- nity, particularly human immunodeficiency virus (HIV) in- [1, 2]. In sub-saharan Africa, C. neoformans has become the most frequent cause of adult meningitis in the last decade [3] (Heyderman R, et al., unpublished data). Received 3 February 1997; revised 27 May 1997. It has been suggested that different cytokine profiles are Ethical approval of this study was given by the Medical Research Council of Zimbabwe. reflected in differences in the clinical presentations of bacterial, Reprints or correspondence: Dr. I. M. Hoepelman, University Hospital viral, or tuberculous meningitis. Acute bacterial meningitis, Utrecht, Dept. of Internal Medicine, Division of Infectious Diseases and AIDS, with its clinical picture of abrupt onset, severe illness, and high P.O. Box 85500, 3508 GA Utrecht, The Netherlands. mortality, is characterized by a sharp increase in the proin- The Journal of Infectious Diseases 1997;176:1633 6 flammatory cytokines tumor necrosis factor (TNF)-a, interleu- 1997 by The University of Chicago. All rights reserved. 0022 1899/97/7606 0033$02.00 kin (IL)-1b, IL-6, and IL-8 [4]. In contrast, the clinically milder