Since the mid-1990s, the use of HAART has spread, and the AIDS mortality rate has decreased by more than 80% in the industrialized world . HIV infection/AIDS is becoming a controllable chronic infection and HIV-infected individuals are now living longer. Many HIV-positive people are getting married and wishing to have children.
Semprini et al.  reported that they had conducted artificial inseminations in more than 2000 HIV-discordant couples (HIV-positive male and HIV-negative female) using their swim-up method, and that no HIV transmission was observed. However, their method may be suboptimal because it has not been proven to remove HIV RNA completely, and they did not measure proviral DNA in infected cells in the semen. Zhang et al. reported that HIV may be present as proviral DNA in seminal cells in HIV-infected men who have achieved undetectable levels of viral RNA in plasma with HAART, and this HIV could be capable of sexual transmission. It has not been determined whether HIV is attached to spermatozoa or whether spermatozoa can be infected with HIV [4,5]. Therefore, contraception is recommended for HIV-discordant couples, even if HIV RNA is undetectable in plasma .
Authorities in different countries have different opinions concerning the use of assisted reproductive technology using spermatozoa collected by the swim-up method [6–8]. However, it would be possible for an HIV-infected man to father children without risk of HIV transmission if HIV-free spermatozoa can be obtained from his semen. This study examines an improved swim-up method for isolating HIV-free sperm and its use in assisted reproductive methods.
This clinical study was approved by the ethics committees of Niigata University, Ogikubo Hospital, Keio University and Kyorin University. All of the couples visited the Hematology Department of Ogikubo Hospital and received counselling and explanations of the clinical study. Informed consent was obtained from all participating couples. Semen was obtained by masturbation, and then tested for sperm concentration, motility and deformity.
An isotonic solution of Percoll (Amersham Life Science, Tokyo, Japan) was made by dissolving 980 g Percoll in 10.0 ml 2.0 mol/l Hepes-NaOH, pH 7.4, 10.0 ml human serum albumin (25%w/v), 0.05 g fosfomycin and 0.05 g cefarotin. The resulting 98% Percoll solution was sterilized with a Millipore filter (0.45 μm pore size).
The procedure is shown in Fig. 1. Ejaculates were diluted twice with Hanks solution, followed by standing in a test tube for 10 min to precipitate filterable micro-calculus, then filtered through an ART filter (20 μm clearance; ART filter, Nipro, Osaka, Japan) to remove fibers, micro-calculus and mucinous debris. The upper phase of sperm suspension was loaded onto 6 ml Percoll linear gradient from 98% to 30% in a separable fine-neck tube (Nipro) and centrifuged at 400 × g for 30 min. The separable fine-neck tube was made of glass, and its bottom was squeezed to minimize the volume of sediment. To recover the sperm precipitated in the bottom tip, the top of the tube was plugged with a rubber cap, and the middle of the squeezed bottom was snapped off with an ampoule cutter.
Motile sperm were separated by the modified swim-up method. A fine glass capillary was inserted in 2 ml of the medium in a vial, then a needle tip was introduced to the bottom through the inner capillary. The motile sperm were allowed to swim up at 37°C in an incubator with 5% CO2–air. After 60 min, 1 ml of upper layer was collected, containing the sperm that had swum up.
The sperm suspension was divided into two portions. One was used for HIV assessment, and the other was cryopreserved with KS-II medium  in a liquid nitrogen container.
Standard HIV-1 materials
MOLT-4 cells infected with HIVLAI and its culture supernatant were used as standards for HIV-1-infected cells and virus stock, respectively. RNA purified from virus stock and the pNL4-3 plasmid  were the standards for HIV-1 RNA and DNA, respectively. The concentrations of the standard HIV-1 DNA and RNA were determined by spectrophotometry and the null-class equation of the Poisson distribution of the reverse transcriptase (RT)-nested polymerase chain reaction (PCR). Cells were counted using a Burker–Turk hemocytometer (Emergo, Landsmeer, the Netherlands). The virion concentration was considered to be half the virus RNA concentration.
Detection of HIV-1 RNA and DNA
The samples of sperm suspension, culture medium or plasma were centrifuged at 35 500 × g for 1 h at 4°C. RNA and DNA were extracted from the precipitate using QIAamp UltraSens Virus Kit (Qiagen, Tokyo, Japan). One fourth of the eluate was tested in quadruplicate by RT-nested PCR as follows. The RT reaction was performed by incubation at 42°C for 10 min in a 20 μl solution consisting of 1× PCR buffer II (10 mmol/l Tris-HCl pH 8.3, 50 mmol/l KCl; Perkin Elmer Life Sciences, Yokohama, Japan), 3 mmol/l MgCl2, 0.2 mmol/l each dNTP, 0.1 μmol/l primer GA1R (5′-CCCAGGATTATCCATCTTTTATAG-3′, 1595–1572 ), 4 U RNasin (Promega, Tokyo) and 20 U SuperScript II (Invitrogen, Tokyo, Japan). The whole RT product was subjected to a first-round PCR in a 50 μl solution consisting of 1× PCR buffer II, 4 mmol/l MgCl2, 0.2 mmol/l each dNTP, 0.2 μmol/l primers GA1F (5′-TGTTAAAAGAGACCATCAATGAGG-3′, 1388–1411) and GA1R and 0.5 U AmpliTaq (Perkin-Elmer). Then, 1 μl of the first-round PCR product was used in the second-round PCR in a 50 μl solution containing primers GA2F (5′-GGCCAGATGAGAGAACCAAGG-3′, 1465–1485) and GA2R (5′-CATCCTATTTGTTCCTGAAGGGTAC-3′, 1535–1511) and the other components in first-round PCR. The primers were located in gag p24. The thermal profile of PCR in GeneAmp PCR System 9700 (Applied Biosystems, Tokyo, Japan) was 94°C for 2 min; three cycles of 94°C for 5 s,48°C for 10 s and 72°C for 15 s; 22 cycles of 94°C for 5 s, 60°C for 10 s and 72°C for 15 s; with a final cycle of 72°C for 1 min and then the mixture kept at 4°C. The PCR products were electrophoresed through a 2.0% agarose gel in the presence of 0.5 μg/ml ethidium bromide and photographed under ultraviolet illumination. Throughout the procedure, the medium used for washed sperm or fertilized eggs was the negative control and this medium with 10 virions added was the positive control. The whole process took approximately 5 h. For samples of peripheral blood mononuclear cells (PBMC), DNA was extracted using QIAamp DNA Kit (Qiagen) and 0.5 μg of the DNA was tested in triplicate by the PCR procedures omitting reverse transcription. Competitive RT-nested PCR was performed as previously described .
Infectivity of HIV-1 during incubation
After incubation at 37°C under 5% CO2 for various periods, the virus stock was added to 5 × 106 stimulated donor PBMC in 1 ml RPMI 1640 medium supplemented with 30% immobilized fetal calf serum and 70 U/ml human recombinant interleukin 2 (Shionogi, Osaka, Japan), and further incubated for 5 days. The culture supernatants were tested for p24 concentration with VIDAS HIV P24 II (BioMérieux, Tokyo, Japan).
If the HIV-1 testing for virion RNA and proviral DNA was negative, the other portion of frozen sperm was thawed for use in assisted reproduction. Mature eggs were obtained by means of ovulation-inducing drugs, and then placed in a dish containing 3 ml RPMI culture medium (20% albumin). The HIV-1–negative sperm solution was introduced to eggs by means of in vitro fertilization (IVF) or intracytoplasmic sperm injection (ICSI), and the dish containing treated eggs was incubated at 37°C under 5% CO2 for 48 h. Before embryo transfer, the culture medium for the fertilized egg was collected and tested for the presence of HIV-1 again. Only when HIV-1 RNA and DNA assays by nested PCR were negative was embryo transfer conducted. All the female partners who underwent assisted reproductive therapy, even those who did not conceive successfully, were tested for HIV antibodies, HIV-1 RNA and proviral DNA in the blood at 1 and 3 months after the assisted reproductive technique and after delivery. The babies were tested for HIV RNA and proviral DNA in umbilical cord blood at birth and in blood until 6 months after birth.
Sensitivity of the HIV-1 RNA/DNA test
The procedure to detect a single copy of either HIV-1 virion RNA or proviral DNA in sperm suspensions (the HIV-1 RNA/DNA test) was developed by selecting and improving techniques in three main steps (collection of HIV-1 virions and infected cells by centrifugation, extraction of viral RNA and DNA with silica-gel-membrane technology, and the detection of the viral RNA and DNA by nested PCR) to achieve zero apparent loss in recovery at each step. First, the exact virion concentration of the standard HIVLAI virus stock was determined by direct RT-nested PCR at endpoint dilution by using the null-class equation of the Poisson distribution. Then, one virion of HIVLAI, on average, was added to 1 ml Sydney IVF medium (Cook, Tokyo, Japan) and the whole procedure was initiated. When one fourth of the eluate from an extraction column was examined (replicated four times) with RT-nested PCR, 12 of 20 samples exhibited at least one band in four reactions (Fig. 2). Next, a single MOLT-4 cell chronically infected with HIVLAI was added to 1 ml Sydney IVF medium and subjected to the HIV-1 RNA/DNA test without reverse transcription. RT-nested PCR showed that 6 of 10 samples exhibited positive reactions. The ratios of positive reactions for virions (60%) and infected cells (60%) were in close agreement with that predicted from the Poisson distribution (63%), providing evidence that the protocol has the ability to detect RNA/DNA in a single virion as well as in a single infected cell when present in as much as 1 ml of IVF medium. To study the influence of the presence of sperm in the medium on the sensitivity of the test, two sets of five samples containing 0.5, 1, 2, 4 and 8 × 106/ml spermatozoa in Sydney IVF medium were tested; one set was mixed with 50 virions and the other set with 100 infected cells. The numbers of virion RNA and proviral DNA from sperm-containing samples that were determined by competitive PCR varied in the range 75−112 copies (note two RNA copies/virion) and 96−122 copies, respectively, in a manner that was not dependent on the sperm quantity. These results strongly suggest that the protocol can detect a single virion or infected cell even in the presence of up to 8 × 106 spermatozoa per sample.
Removal of HIV-1 virions and infected cells from mixed semen by sperm-washing
To assess the efficiency of sperm-washing procedures with Percoll density gradient centrifugation and swim-up for removal of HIV-1 from semen, HIV-1 virions or HIV-1–infected cells were added to healthy donor semen. When 2 × 107 virions HIV-1 were mixed with 1.6 ml healthy donor semen containing 6.3 × 107 spermatozoa/ml, 63 copies of HIV-1 RNA were detected after centrifugation but no HIV-1 RNA was detected after swim-up. When 5 × 105 HIV-1–infected cells were mixed with 1.6 ml of healthy donor semen containing 6.3 × 107 spermatozoa/ml, no HIV-1 DNA was detected after either centrifugation or swim-up. The sperm suspension collected after swim-up was 1.0 ml in volume and contained 50 000 spermatozoa of 100% motility.
Decay of infectivity of HIV-1 during incubation
A virus solution of HIVLAI was incubated in culture medium for various periods and the p24 production ability was quantified in stimulated PBMC to evaluate the stability of HIV-1 in vitro with regard to infectivity. Infectivity decreased semiexponentially with a half-life of approximately 13 h.
Results of the clinical study
A total of 52 HIV-1-positive individuals participated in the clinical study (Table 1); 29 were haemophiliacs and 23 had become infected through sexual contact. The median age was 33 years (range, 27–44) in the 16 untreated individuals, 34 years (range, 28–41) in patients receiving antiretroviral drugs and with viral load ≥ 50 copies/ml, and 32 years (range, 20–51) in patients receiving HAART and with viral load < 50 copies/ml. Median plasma viral load was 17 500 copies/ml (range, 70–100 000) in the untreated group and 1500 copies/ml (range, 54–31 000) in patients receiving treatment and with a viral load ≥ 50 copies/ml.
Among 48 patients whose partner had assisted reproductive therapy, the median plasma viral load was 17 500 copies/ml (range, 70–100 000) in 15 patients in the untreated group, 4800 copies/ml (range, 54–31 000) in 10 patients receiving antiretroviral treatment and with viral load ≥ 50 copies/ml, and < 50 copies/ml in 23 patients taking HAART. Median CD4 cell count was 365 cells/μl (range, 66–1071) in the untreated group, 457 cells/μl (range, 60–652) in patients receiving antiretroviral drugs and with viral load ≥ 50 copies/ml, and 399 cells/μl (range, 41–792) in patients receiving HAART and with viral load < 50 copies/ml. The median sperm count of the HIV-positive males was 47 × 106/ml (range, 0–82) in the untreated group, 41 × 106/ml (range, 0–65) in patients receiving antiretroviral drugs and with viral load ≥ 50 copies/ml, and 35 × 106/ml (range, 0–120) in patients receiving HAART and with viral load < 50 copies/ml.
Azoospermia occurred in four patients, who were excluded from this study.
In all patients, the median motility rate was 28%, and the median incidence of morphologically normal spermatozoa was 12%. The median concentration of spermatozoa in patients (excluding the four patients with azoospermia) was 42 × 106/ml (range, 3–120) and 52 × 106/ml (range, 0–170) spermatozoa were collected after the Percoll centrifugation. The median motility rate was 28% and 45% before and after the Percoll procedure. Following the swim-up method, there were 1.5 × 106/ml (range, 0–11) collected spermatozoa, and the motility rate was 100%. Spermatozoa could be collected by the swim-up method in 73 semen samples from the 48 patients. No HIV-1 RNA or proviral DNA was detected in any sperm suspensions collected after the swim-up procedure. The HIV-1-negative sperm was used for IVF in 12 couples and for ICSI in 31 couples. HIV-1 RNA or proviral DNA could not be deptected in the culture medium of the fertilized eggs before embryo transfer. Of the 43 female partners, 20 conceived and 27 babies were born. HIV antibodies, HIV RNA and proviral DNA were negative in all of the females and babies.
This study demonstrated that it is possible to detect a single copy of HIV-1 RNA or proviral DNA, and that HIV-negative spermatozoa can be obtained from the semen of HIV-positive males with the careful use of density gradient centrifugation and the swim-up technique. There has been no HIV-1 transmission in any of the female partners who underwent IVF or ICSI, nor in any of the babies.
Some studies have indicated that HIV can bind and enter into spermatozoa [4,5,12,13]. However, CD4 is not expressed on the surface of spermatocytes or spermatozoa [14,15]. Brogi et al.  have reported that HIV can attach to the surface glycoprotein of spermatozoa. In children at birth, the infection route is considered to be mother to child , and there is no case report of a child or embryo who has been infected with HIV via spermatozoa. It has also not been proven that a spermatid could be infected with HIV during spermatogenesis. This study showed that spermatozoa collected by the swim-up method were neither infected with HIV-1 nor had HIV-1 attached to them.
Semen contains spermatozoa, seminal plasma, white blood cells, microbes, metallic crystals and fibres of underwear. If components with higher density than spermatozoa are in a sample at centrifugation, those components may bring viruses and infected cells down to the bottom sperm fraction. Therefore, in our technique, we left diluted semen undisturbed to settle heavy components, and then took the sperm-containing upper fraction. If the sperm fraction (the bottom layer) following Percoll centrifugation is pipetted through the other denser layers, as is commonly done, HIV may contaminate the sperm fraction via the tube wall. In this study, we sealed the top of the tube after centrifugation and collected the sperm fraction by cutting off the bottom layer, which prevented contamination from the higher layers.
Gomibuchi et al.  reported that their method could not reduce HIV-1 RNA in semen to < 100 copies/ml in 55.6% of patients. Kuji et al.  have reported that the use of endotoxin-free Pureseption for semen processing had a lower elimination rate for HIV than the Percoll method. Some groups have used a swim-up technique in which the spermatozoa collected after centrifugation with a separating solution were washed with a culture medium and layered below the medium, followed by swim-up. Because the difference in the specific gravity of the sperm suspension and that of the culture medium is small, HIV and mononuclear cells may easily diffuse to the top layer during the swim-up method . The actual procedures of the swim-up method, such as semen-washing techniques, the materials used in centrifugation, the concentrations of separating solutions, and the methods used to collect the bottom layer (sperm fraction) vary among researchers [17,19,20]. Therefore, it is considered that the HIV elimination rate will also vary. Our improved swim-up method provides a safer procedure for use in assisted reproductive techniques.
Semprini et al.  have reported that HIV transmission has not occurred in over 2000 patients who underwent artificial insemination using their method. Their successful results may be explained by the fact that infectious HIV is less than 1/10 000 of all HIV virions [11,21,22] and that removal of the HIV-producing mononuclear cells by the swim-up method is a major factor in reducing infection risk. We have reported that a female was infected with HIV-1 after six artificial insemination procedures using sperm prepared only by centrifugation in another hospital . Artificial insemination should not be performed when inadequate HIV elimination methods are used or when the absence of HIV is not confirmed by highly sensitive tests.
Most HIV-infected patients in this study had low sperm counts and sperm motility rates, and provided a small number of spermatozoa after the swim-up method. As we try to achieve higher virus elimination rates, the number of collected spermatozoa becomes small. Ohl et al.  reported no pregnancies after artificial insemination using sperm obtained by the swim-up method. If it takes too long for PCR procedures, or if spermatozoa are frozen, the fertilization ability of the spermatozoa may be decreased and the probability of pregnancy may be low. It is difficult to confirm rapidly the removal of HIV-1 RNA and DNA in spermatozoa actually used for artificial insemination. CD4 and chemokine receptors are not expressed on eggs  and, therefore, eggs cannot become infected with HIV in the sperm suspensions collected using the swim-up method even if HIV is present in the suspension. If the suspensions are contaminated with a small amount of HIV, the infectivity of the HIV would still decrease to below 1/10 after a 2 day incubation. In addition, in IVF or ICSI, it is possible to confirm the absence of HIV-1 in the culture medium of fertilized eggs before embryo transfer. Therefore, we conducted IVF or ICSI using frozen spermatozoa that had been confirmed negative for HIV-1.
In conclusion, we have demonstrated that it is possible to collect spermatozoa with evidence of the absence of HIV-1 RNA and proviral DNA from semen of HIV-infected males. Whatever method is used for assisted reproductive technique and for removal of HIV from semen to reduce the risk of secondary transmission, it is essential to confirm the absence of HIV-1 RNA and proviral DNA in the sperm preparation used for the assisted reproductive technique with the most sensitive tests possible.
1. Hammer SM, Squires KE, Hughes MD, Grimes JM, Demeter LM, Currier JS, et al
. A controlled trial of two nucleoside analogues plus indinavir in persons with human immunodeficiency virus infection and CD4 cell counts of 200 per cubic millimeter or less. AIDS Clinical Trials Group 320 Study Team. N Engl J Med 1997; 337:725–733.
2. Semprini AE, Fiore S, Pardi G. Reproductive counselling for HIV-discordant couples. Lancet 1997; 349:1401–1402.
3. Zhang H, Dornadula G, Beumont M, Livornese L Jr, van Uitert B, Henning K, Pomerantz RJ. Human immunodeficiency virus type 1 in the semen of men receiving highly active antiretroviral therapy. N Engl J Med 1998; 339:1803–1809.
4. Brogi A, Presentini R, Moretti E, Strazza M, Piomboni P, Costantino-Ceccarini E. New insights into the interaction between the gp120 and the HIV receptor in human sperm (human.sperm/gp120/galactoglycerolipid/antigalactosylceramide/seminolipid/spermatogonia). J Reprod Immunol 1998; 41:213–231.
5. Bagasra O, Farzadegan H, Seshamma T, Oakes JW, Saah A, Pomerantz RJ. Detection of HIV-1 proviral DNA in sperm from HIV-1-infected men. AIDS 1994; 8:1669–1674.
6. Garrido N, Meseguer M, Simon C, Pellicer A, Remohi J. Assisted reproduction in HIV and HCV infected men of serodiscordant couples. Arch Androl 2004; 50:105–111.
7. Weigel MM, Gentili M, Beichert M, Friese K, Sonnenberg-Schwan U. Reproductive assistance to HIV-discordant couples: the German approach. Eur J Med Res 2001; 6:259–262.
8. Centers for Disease Control and Prevention. Epidemiologic Notes and Reports HIV-1 Infection and Artificial Insemination with Processed Semen.MMWR
9. Kaneko S, Kobayashi T, Lee HK, Won WK, Oda T, Izumi Y, et al
. Cryogenic preservation of low-quality human semen. Arch Androl 1990; 24:81–86.
10. Adachi A, Gendelman HE, Koenig S, Folks T, Willey R, Rabson A, et al
. Production of acquired immunodeficiency syndrome-associated retrovirus in human and nonhuman cells transfected with an infectious molecular clone. J Virol 1986; 59:284–291.
11. Kato S, Hiraishi Y, Nishimura N, Sugita T, Tomihama MM, Takano T. A plaque hybridization assay for quantifying and cloning infectious human immunodeficiency virus type 1 virions. J Virol Meth 1998; 72:1–7.
12. Nuovo GJ, Becker J, Simsir A, Margiotta M, Khalife G, Shevchuk M. HIV-1 nucleic acids localize to the spermatogonia and their progeny. A study by polymerase chain reaction in situ hybridization. Am J Pathol 1994; 144:1142–1148.
13. Baccetti B, Benedetto A, Burrini AG, Collodel G, Ceccarini EC, Crisa N, et al
. HIV-particles in spermatozoa of patients with AIDS and their transfer into the oocyte. J Cell Biol 1994; 127:903–914.
14. Gil T, Castilla JA, Hortas ML, Molina J, Redondo M, Samaniego F, et al
. CD4+ cells in human ejaculates. Hum Reprod 1995; 10:2923–2927.
15. Kim LU, Johnson MR, Barton S, Nelson MR, Sontag G, Smith JR, et al
. Evaluation of sperm washing as a potential method of reducing HIV transmission in HIV-discordant couples wishing to have children. AIDS 1999; 13:645–651.
16. Peckham C, Gibb D. Mother-to-child transmission of the human immunodeficiency virus. N Engl J Med 1995; 333:298–302.
17. Gomibuchi H, Nagamatsu A, Minoura S, Nonoyama M, Tachikawa N, Oka S. Intrauterine insemination with processed semen of HIV positive husband: study ofsemen processing. J AIDS Res 2001; 3:6–9.
18. Kuji N, Yoshii T, Kato S, Asada H, Sueoka K, Yoshimura Y. Sedimentation kinetics of HIV-1 in two gradient media.58th Annual Meeting of American Society for Reproductive Medicine
. Seattle, 2002, pp. 12–17.
19. Meseguer M, Garrido N, Gimeno C, Remohi J, Simon C, Pellicer A. Comparison of polymerase chain reaction-dependent methods for determining the presence of human immunodeficiency virus and hepatitis C virus in washed sperm. Fertil Steril 2002; 78:1199–1202.
20. Hanabusa H, Kuji N, Kato S, Tagami H, Kaneko S, Tanaka H, et al
. An evaluation of semen processing methods for eliminating HIV-1. AIDS 2000; 14:1611–1616.
21. Layne SP, Merges MJ, Dembo M, Spouge JL, Conley SR, Moore JP, et al
. Factors underlying spontaneous inactivation and susceptibility to neutralization of human immunodeficiency virus. Virology 1992; 189:695–714.
22. McKeating JA, McKnight A, Moore JP. Differential loss of envelope glycoprotein gp120 from virions of human immunodeficiency virus type 1 isolates: effects on infectivity and neutralization. J Virol 1991; 65:852–860.
23. Hanabusa H, Kato S, Kaneko S, Suzuki M, Takakuwa K, Kuji N, et al
. Clinical outcome of artificial reproductive technology using HIV-1 free sperm and the problems.17th Congress of the Japanese Society for AIDS Research
. Kobe, November 2003 [abstract 148].
24. Ohl J, Partisani M, Wittemer C, Schmitt MP, Cranz C, Stoll-Keller F, et al
. Assisted reproduction techniques for HIV serodiscordant couples: 18 months of experience. Hum Reprod 2003; 18:1244–1249.
25. Baccetti B, Benedetto A, Collodel G, Crisa N, di Caro A, Garbuglia AR, et al
. Failure of HIV-1 to infect human oocytes directly. J Acquir Immune Defic Syndr 1999; 21:355–361.