Abstract
This study investigated the prevalence of adeno-associated virus (AAV) and human papillomavirus (HPV) DNA in cervical samples of HIV-seropositive and -seronegative women attending a clinic in south-eastern Brazil. Both viruses were investigated by polymerase chain reaction (PCR) and cytological exams were performed. AAV was typed by PCR and restriction fragment length polymorphism analysis. AAV prevalence was 19.7% (56/284), with 18.7% (21/112) and 20.3% (35/172) in HIV-positive and -negative women, respectively. AAV type 2 was the single virus type detected. AAV was detected with higher frequency in HPV-infected women (P < 0.05) as was HPV in HIV-positive women (P < 0.05). The AAV–HPV co-infected women showed a lower rate of atypical squamous cells of undetermined significance or cervical intraepithelial neoplasia development compared with those infected only with HPV. The prevalence of AAV2 confirms this type as the most common in human samples. This is the first report examining AAV in cervical samples of HIV-infected women and indicates that HIV infection does not appear to influence AAV prevalence or AAV–HPV co-infection.
Keywords
INTRODUCTION
Adeno-associated virus (AAV) belongs to the Parvoviridae family, Dependovirus genus and is composed of 12 species named AAV1–AAV12 that are known to infect primates. 1–6 AAV2, AAV3, AAV5 and AAV9 have been reported in human samples such as blood, endometrial biopsy, cervical secretion, semen and fragments from spontaneous miscarriage and testicular biopsy. 7–13 AAV2 is the most frequently detected species and is considered to be the prototype. 14–16
AAVs are rather unique viruses since they need a non-related helper virus such as an adenovirus, human herpesvirus or a human papillomavirus (HPV) to cause a productive infection in humans. 9,10,17 Co-infection with HPV, more frequent than with other helper viruses, was described in cervical epithelium. 9,10,18 A possible bidirectional action was also observed, with AAV acting as a suppressor agent on HPV replication. 19–22 It becomes more meaningful if one takes into account the fact that HIV favours HPV infection and increases the persistence of HPV and consequently the progression of cervical lesions to cervical carcinoma. 23–25 AAVs, on the other hand, may play a protective role in HIV-positive women intervening in the course of HPV infection. The prevalence of AAV in HIV-positive women, infected or not with HPV, has not yet been reported.
This study aimed to determine the prevalence of AAV and HPV in HIV-positive and -negative women, who were attending a sexually transmitted disease (STD) centre and to describe the interaction between these viruses and cervical lesions.
MATERIALS AND METHODS
Cases and samples
Two hundred and eighty-four women who were attending the Reference Center of STD, Vitória, south-eastern Brazil, from March to December 2006 partook in this study. After providing written informed consent, samples of endocervical secretions were collected from the participants with a cytobrush and maintained at −70°C in TE buffer 1× (Tris 10 mmol/L, EDTA 1 mmol/L – pH 7.0) until DNA extraction. The HIV serology and cytopathological evaluation was conducted by the Public Central Laboratory of Vitória Municipality. This study was approved by the Ethical Committee in Research at the Federal University of Espírito Santo, Brazil.
AAV and HPV detection
DNA was extracted using a QIAamp® DNA Mini Kit (Qiagen, Valencia, CA, USA), according to the manufacturer's instructions; the extraction and DNA integrity were controlled with β-globin gene detection by polymerase chain reaction (PCR). 26
AAV was investigated by first round PCR and nested-PCR with pairs of primers Pan1/Pan3 and Nest1/Nest2, as described previously, 8 which amplify the rep gene region of AAV2, AAV3 and AAV5, resulting in amplicons of 337 and 158 bp, respectively. HPV DNA was investigated by PCR, using the protocol previously established, 27 which detects a fragment of 450 bp of the L1 gene region with consensus primers MY09/11. 28 Positive controls for AAV2 and HPV-16 were kindly supplied by Dr Jörg R Schlehofer (Virology Laboratory, Heidelberg, Germany) and Dr Ledy HS Oliveira (Fluminense Federal University, Rio de Janeiro, Brazil), respectively.
AAV molecular characterization
AAV5 was discriminated from the AAV-positive cases by a PCR and hemi-nested-PCR with pairs of primers A5/Pan3 and A5/Nest2. 8
A protocol of restriction fragment length polymorphism (RFLP) was established in this study using five units of HaeIII and Sau3AI restriction enzymes to discriminate AAV types 2, 3 and 5. Three micrograms of the 158-bp nested-PCR products were digested in separate reactions during overnight incubation and the fragments were resolved in a 3% agarose gel. HaeIII and Sau3AI generate fragments of digestion from AAV2 (123 and 35 bp) and AAV3 types (104, 36 and 18 bp), respectively.
Statistical analysis
Standard descriptive statistical analyses were performed, including frequency distribution for categorical data and calculation of means and interquartile range for continuous variables. Prevalence was calculated to reflect the cumulative frequency of outcome, with corresponding 95% confidence intervals (CI). Odds ratio (OR) and 95% CI were calculated in bivariate analysis to estimate the strength of the association between outcome and each potential risk factor.
RESULTS
A total of 284 women were investigated, with a median age of 30 years (range 18–49). HIV serology identified 112 seropositive and 172 seronegative women. Cervical findings were: 263 with normal cytology, 10 with atypical squamous cells of undetermined significance (ASCUS) and 11 cases of cervical intraepithelial neoplasia class I and II (CIN I/II).
To assess the prevalence of AAV and HPV in cervical scrapings of HIV-positive and -negative women, specific sequences of both viruses were amplified. Cellular DNA was detected in all cervical samples (β-globin sequence).
HPV DNA was detected in 133 samples (47%): 56% (63/112) in HIV-positive and 40.7% (70/172) in HIV-negative women (P < 0.05). AAV total prevalence was 19.7% (95% CI 11.3–28.1%), with 18.7% (21/112) in HIV-positive and 20.3% (35/172) in HIV-negative women (P > 0.05).
The RFLP method revealed the same digestion pattern (123/35 bp) for all samples corresponding to AAV2. Products of amplification of AAV5 were not observed in the specific PCR.
AAV DNA was detected in 36 of the 133 HPV-positive samples (27%) but only 20 of the 151 HPV-negative samples (13%), a significant correlation (OR 2.43, CI 1.33–4.46). AAV–HPV co-infection was observed in 16% (18/112) and in 10% (18/172) of HIV-positive and HIV-negative women (P > 0.05), respectively.
Distribution of AAV results in HPV-positive women cases according to cervical cytological findings, observed in cervical samples from HIV-positive and -negative women attending the Reference Center of STD/AIDS, Vitória-ES, south-eastern Brazil
AAV = adeno-associated virus; HPV = human papillomavirus; CIN = cervical intraepithelial neoplasia; ASCUS = atypical squamous cells of undetermined significance; OR = odds ratio; CI = confidence interval
*Viral combinations based on AAV positivity in HPV-infected women
†Total n = 284
‡Total of cases (including AAV+ and AAV−/HPV+ and HPV −): CIN = 11; ASCUS = 10; normal = 263
Sociodemographic characteristics of AAV DNA-positive and -negative women
AAV = adeno-associated virus; SD = standard deviation; IQR = interquartile range
DISCUSSION
AAV, a sexually-transmitted parvovirus without a definitive association to any human disease thus far, exerts a suppressive role on transformation induced by oncogenic viruses, such as HPV. 22,29,30 As a satellite virus, AAV requires co-infection with a helper virus to cause productive infection and HPV is reported to be the most prevalent such virus in the AAV-infected genital tract. 9,10,18
Data show that HPV is associated with a greater persistence and evolution of cervical lesions in HIV-infected women than in non-infected women. 31–33 However, there is no study investigating AAV in HIV-infected women and rates of co-infection among these three viruses is unknown. Hence, this investigation explored AAV and HPV infections in women infected or not with HIV.
In this study HPV prevalence was higher among HIV-positive women than in seronegative women (56% versus 40.7%), corroborating previous studies worldwide. 34–36 AAV total prevalence was also similar to previous reports. 15,37,38 HIV infection did not seem to effect AAV infection or AAV–HPV co-infection rates. However, AAV was significantly associated with HPV-infected women with a risk around two times higher in HPV-positive samples. During primary infection, in the absence of helper virus, AAV establishes latent infection and re-activates when superinfection or subsequent infection occurs with a helper virus. 39 This fact might explain the higher prevalence of AAV observed in the HPV-infected women and previously described in genital samples. 9,10,18
The bidirectional interaction between AAV–HPV is very complex. AAV has been demonstrated to inhibit in vitro the oncogenic transformation caused by bovine papillomavirus (BPV) and by HPV-16 and -18, mediated by AAV-encoded Rep78 protein. 22,29,30,40 Meyers et al. 20 suggested that when AAV infects a cell culture system, at low multiplicity of infection (MOI), it mildly enhances HPV replication, whereas, at high MOI, it inhibits HPV replication. Hermonat et al. 41 reported that BPV suppression is increased when the interval of infection between the AAV and BPV is shorter. Thus, AAV might have a protective role on the oncogenic transformation caused by papillomavirus in a dose-dependent or temporal infection manner.
Few women investigated showed cytological abnormalities. Even though, the statistical regression analysis (OR association) showed that AAV infection reduced the chance of cervical lesions (ASCUS/CIN) in HPV-positive women. This fact was not observed in the AAV-negative cases. Certainly, a larger number of samples will reinforce the findings of the possible protective role of AAV on HPV-induced lesions. This result supports previous studies that described an inhibitory action of AAV in HPV-induced lesions. 42–44 Conversely, some studies do not agree with this inverse association between AAV and cervical tumorigenesis. 45–47 Hence, the AAV–HPV co-infection should be further investigated to highlight the protective association on cervical malignant progression.
The hypothesis that AAV may confer a protective role in HIV-positive women by intervening in the course of HPV infection was not observed in the present study. However a larger number of samples may be needed to clarify this matter.
Limitations of the present study include the descriptive design precluding inferences on cause–effect and the modest sample size. However, the response rate was high and the population bases of the data are very relevant and can be generalized to women in the STD clinic.
In conclusion, AAV prevalence is influenced by HPV infection and HIV does not seem to have any effect on AAV detection, in spite of its enhancement of HPV infection. Furthermore, studies following primary infection of AAV and/or HPV may elucidate the real association among these sexually transmitted viruses and any protective action of AAV on malignant transformation induced by HPV.
Footnotes
Acknowledgements
This work was supported by Fundação de Apoio à Ciência e Tecnologia do Espírito Santo grants (FACITEC). We thank Dr Jörg R Schlehofer who supplied the positive control of AAV and Dr Ledy HS. Oliveira for the positive control of HPV-16.