[ Next ] [ Prev ] [ Abs ] [ Chi ] [ TOC ] [ Home ]
Chin Med J (Taipei) 1997;59:325-33.
1Department of Medicine, Taipei Medical College; 2National Institute of Preventive Medicine, Department of Health, Executive Yuan, Taipei, Taiwan, R.O.C.
Background. This study is designed to resolve the problem of whether temperature or freeze/thaw cycle will have any impact on the sensitivity for detection of anti-HIV-1 antibody by particle agglutination (PA), enzyme-linked immunosorbent assay (ELISA) and western blotting (WB). To reduce potential risk for laboratory personnel exposed to HIV-infection, it will be useful to determine the temperature effect on HIV infectivity.
Methods. Testing sera were incubated at different temperatures or treated with several cycles of freeze and thaw. PA, ELISA and WB were used to detect anti-HIV-antibodies, whereas syncytia formation assay and polymerase chain reaction (PCR) were applied to detect HIV-infection.
Results. The data showed that certain temperature points (no treatment, 25OC for 1hr, 2hrs and 4 hrs, 37OC for 30 minutes and 60 minutes, 56OC for 30 minutes and 60 minutes, 65OC for 15 minutes and 30 minutes) had no impact on the testing results of ELISA, PA and WB in detection of anti-HIV-1 antibody. In addition, testing results of 50 normal human serum samples which had been heated to 56OC for 30 minutes were still negative by ELISA and PA. Only the samples incubated at 65OC for 60 minutes had slight differences in results. Freeze and thaw treatments of the serum did not alter anti-HIV testing results, either. Treatments of supernatant of HTLV-IIIB culture at 56OC for 30 minutes and 60 minutes, 65OC for 15 minutes and 30 minutes could eliminate the syncytia formation caused by HIV-infection. Further analysis of the samples by PCR was able to detect HIV-specific sequences in all the treatments.
Conclusions. Anti-HIV antibody is quite stable in serum, even when it is pre-heated to 56OC for 30 minutes. Freeze and thaw treatment of serum samples up to seven cycles did not change the results, either. In addition, to minimize the potential risk of laboratory personnel exposed to HIV infection, pre-treatment of serum samples with heat at 56OC for 30 minutes or 60 minutes can reduce HIV infectivity. However, laboratories still must emphasize the importance of universal precautions rather than heat-inactivation of serum to prevent occupational transmission of HIV.
[Chin Med J (Taipei) 1997;59:325-33.]
Keywords: acquired immunodeficiency syndrome, enzyme-linked immunosorbent assay, human immunodeficiency virus, particle agglutination, western blotting
Received: August 1, 1996.
Accepted: April 21, 1997.
Address reprint requests to: Gong-Ren Wang, 161, Kun Yang St., Nan Kang Dist., Taipei, Taiwan, R.O.C.
Encouraging progress has been made in the understanding the pathogenesis of human immunodeficiency virus (HIV) infection, started from the initial isolation of causative agents in 1983 [1] to the more recent development of potential therapeutic approaches via antiviral drugs and vaccination trials. Nevertheless, the resolution toward eventual control of HIV still seems far away and acquired immunodeficiency syndrome (AIDS) remains untreatable. To date, no drug or vaccine has become available to cure or stop the disease [2,3]. However, the development of sensitive and specific tests for antibody to HIV, the causative agent of AIDS, has progressed rapidly [1]. These tests have been used for various purposes, including clinical diagnosis of HIV infection, for symptomatic and asymptomatic patients who are in counseling and testing programs, for seroprevalence surveys and for blood-donor screening.
Laboratory testing in clinical practice is performed to provide information and guidance to the physician in treating the individual patient. Erroneous reports, positive or negative, may be distressing to the patient and misguide the physician. Testing for antibodies against HIV is especially complicated by the social and medical implications of false reports. A false positive report to a patient is potentially tragic, but largely preventable. A false negative report has serious potential public health significance because the person may then continue to infect others through high-risk sexual or drug-using practices, or by serving as a blood donor; and further available effective medical treatments would not be sought.
Accurate results from diagnosis of HIV infection will assure the safety of blood banks, provide information for physicians to take appropriate steps to treat patients and monitor the epidemiological trend of HIV infection. For example, owing to a shortage of labor, many foreign workers have been imported into Taiwan, and precautionary steps must be taken toward surveillance of any infectious diseases, including HIV infection, they may carry. To monitor the accuracy of HIV-testing results performed by clinicals and hospitals in Taiwan, the Department of Health, Executive Yuan asked the National Institute of Preventive Medicine (NIPM) to implement a systematic proficiency testing program to evaluate laboratory test results. The program has indicated that accurate results could best be determined by sensitivity of the detection kits used and the skills of laboratory assistants. However, some of the participating laboratories complained that the ice needed to maintain stability of tested specimens had melted when samples were received. Thus the concern is raised that unfavorable conditions such as high temperature during the sample delivery process or too many cycles of freeze/thaw in the laboratories probably can deviate the test results.
Currently, in Taiwan, the most popular way to screen HIV-1 infections is either by enzyme linked immunosorbent assay (ELISA) or particle agglutination (PA). Once any positive case has been identified, the infections need to be confirmed by western blotting (WB). This study is designed to resolve the problem of whether the temperature or freeze/thaw cycle will have any impact on the sensitivity of ELISA, PA and WB used to detect anti-HIV-1 antibody [4]. Heat-inactivation is an effective means of destroying HIV-1 and is used to prepare therapeutic blood products [5]. To minimize the potential risk of laboratory personnel's exposure to HIV infection, it was also of interest to find out how temperature affected HIV-1 infectivity. By analysis of this information, determi-nation can be made of the best temperature point at which HIV-1 infectivity is inactivated, yet the integrity of anti-HIV-1 antibody remains sufficiently intact to allow detection by ELISA, PA or WB.
A panel of three serum specimens, kindly provided by the Chinese Blood Service Foundation and previously determined as anti-HIV-1 positive by the methods of ELISA, PA and WB was used as testing samples for evaluation throughout this study. The characteristics of the serum samples are listed on Table 1. According to the results of WB, Samples I and II were weakly positive, whereas Sample III was strongly positive (Figure 1).
Fifty normal human serum samples were obtained for anti-HIV-1 detection by ELISA and PA tests from the Division of Epidemiology of this institute.
Three different types of ELISA kits, manufactured by Abbott, Pasteur, and General Biologicals, among the most commonly applied in Taiwan, were used for evaluation of the testing. PA kit for detection of HIV-1 antibodies was purchased from Fujirebio Inc., Japan (Serodia HIV-1). WB kit for detection of antibodies to HIV-1 was manufactured by Cambridge Biotech, Worcester, MA, USA. The experimental procedures of ELISA, PA and WB followed manufacturers' instructions.
Infectious HIV was measured by determining the ability of heat-treated or freeze/thaw HIV-1 (HXB2 strain) fluids to infect with SupT1 cells. Briefly, cell-free supernatant was serially diluted in individual wells of a 96-well microtiter plate, and immediately 2 x 105 exponentially growing SupT1 cells were added to each well. The wells were examined daily for the presence of syncytia. The first sign of syncytia formation could be seen by 24 hours, with complete syncytia developing by 48 hours and final results read at 96 hours.
In vitro enzymatic amplification was used to identify HIV-1 infection after treatment with different incubation temperatures. Oligonucleotide primers specific for the env and gag were synthesized by Protech Technology enterprise Co., Ltd. Taipei, Taiwan according to the subsequence of HXB2 strain.
The env primers' sequences are 5' AGCAG, CAGGA, AGCAC, TATGG (7347-7366) and 5' CCAGA, CTGTG, AGTTG, CAACAG (7468-7488); gag primers' sequences are 5' ATAAT, CCACC, TATCC, CAGTA, GGAGA, AAT(1097-1124) and 5' TTTGG, TCCTT, GTCTT, ATGTC, CAGAA, TGC (1184-1211). The PCR reaction mixture contained 2 micron g chromosome DNA purified by Genomix Kit (Talent Co, Follateio, Italy), 1.0 micron M each of primers, 200 micron M each of four deoxyribonucleotide triphosphates, 10mM Tris-HCl, pH 7.5, 50mM NaCl, 10 mM MgCl2, and 2.5 unit of Taq polymerase (Promega) in a 100 micron l volume. The amplification was carried out with a DNA thermal cycler (Model 480, Perkin Elmer Cetus, USA). The profile for PCR was as follows: denaturing the samples at 94OC for 2 minutes, annealing at 55OC for 2 minutes, then extension at 72OC for 1 minute totally for 30 cycles. The PCR products were 141 bps and 114 bps in length after amplification with env and gag primers, respectively and they were analyzed by electrophoresis on 3% agarose gels.
HTLV-IIIB/H9 [6] is a HIV-1 (IIIB) permanent productive cell line which was established from the concentrated culture fluids of peripheral blood or bone marrow from several patients with AIDS or related diseases in a cloned permissive neoplastic T cell line H9. This virus appears to be well adapted for in vitro culture in T cells and replicates less in fresh human macrophage.
Non-Hodgkin's T cell lymphoma was isolated from the pleural effusion of an eight-year-old boy, and subcloned on soft agar [7]. High levels of surface CD4 molecules were expressed and can be useful in studies of cell fusion and cytopathic effect (CPE) of HIV infection. They express pan T antigens and lack sheep erythrocytes receptors. SupT1 cells were used for syncytia formation assay as an indicator cell line; they can be maintained in RPMI 1640 plus 10 % fetal bovine serum (FBS).
To evaluate the temperature effect on the stability of anti-HIV antibody, results obtained by ELISA, PA, and WB from samples with different treatments were compared. These treatments included incubations of samples at room temperature (25OC) for 1 hour, 2 hours, and 4 hours, respectively; 37OC for 30 minutes and 60 minutes; 56OC for 30 minutes and 60 minutes; 65OC for 30 minutes and 60 minutes; 75OC for 15 minutes and 30 minutes; 95OC for 5 minutes and 15 minutes. Also the samples were frozen at -70OC, then thawed at 37OC for one, three, five and seven cycles, respectively. When the serum samples were heated to 75OC or above 75OC in this study, they became gel-like in structure and were no longer in a liquid state. Thus only those samples still in liquid state were investigated.
The data obtained by PA are presented in Table 2. It was found that temperature had no effect on the detection of anti-HIV-1 antibody by PA. Even when the serum sample was heated to 65OC for 60 minutes, the results remained the same as for the untreated samples. The freeze and thaw treatment of samples, up to seven cycles, did not change results either. In addition, temperature and freeze/thaw treatments did not change the setting pattern of particles for PA which still showed a defined large ring with a rough multiform outer margin and peripheral agglutination. Additional tests of fifty normal serum samples treated with heat showed no sign of agglutination (data not shown).
Three different manufacturers' ELISA kits were used to detect the anti-HIV-1 antibody (Tables 3-5). Samples I and II had lower optical density (OD) than Sample III. This finding is consistent with the WB result which suggested that Samples I and II are weak positive and Sample III is strong positive.
With reference to the detection kit of General Biological, none of the heating treatments had any apparent effect on the results except at 65OC for 60 minutes of Samples I and II, both of which were weakly positive (Table 3). Treatment at 65OC for 60 minutes did reduce the OD values of Sample I from 0.64 to 0.34 and Sample II from >3 to 1.62, respectively, but the final outcomes were still the same. Over all, increasing temperature did not alter the OD value in any regular patterns. Freeze/ thaw cycles did not change the results either.
For the kit of Abbott (Table 4), most of the heating treatments did not change the OD values except treatment at 65OC for 60 minutes did turn the result of Sample I negative (from 0.165 to 0.098, cut off value: 0.118). OD value of Sample II was reduced from 0.332 to 0.13 (still positive). This treatment had no effect on Sample III (strongly positive). Freeze and thaw treatments had no effect on the three testing samples.
Different heating treatments did not have any effect on the ELISA results obtained by Pasteur (Table 5). The OD values did reduce slightly from 1.7 to 1.39 of Sample I and from 1.79 to 1.63 of Sample II by the treatment at 65OC for 60 minutes, however, both results were still much greater than the cut off value 0.107. Once again, freeze and thaw treatments had no effect on the testing results by Pasteur.
Centers for Disease Control (CDC) have issued a warning about the problems of giving false positive results, problems created by heat-inactivation of serum specimens before HIV-1 antibody testing [4]. Therefore, fifty normal human serum samples were randomly selected for treatment with heat at 56OC for 30 minutes before testing by ELISA (Welcome and Abbott); all of them were negative (data not shown).
The band pattern of sample I was shown in Figure 1A. The p24, gp41 and gp160 remained unchanged throughout the various temperature treatments except at 65OC for 30 minutes. In that condition, the gp160 became almost invisible, but it was still detectable on the blot under the treatment; p41 band could not be seen very clearly on the picture, but was detectable on the blot. The p41 band of Sample II almost disappeared on the picture after the treatment at 65OC for 60 minutes (Figure 1B). However, this band could be seen clearly on the blot after treatment. Bands of p24 and gp160 of Sample II were still very clear on the blot after the treatments. For Sample III, intensity of all the bands, including p15, p24, p31, p51, p55, p66, gp120 and gp160, did not alter after the treatments (Figure 1C). Freeze/thaw treatment of Samples I and III had no effect on the band patterns of WB (Figure 1 D).
The CPEs induced by HIV-infection (syncytia formation) of the treated groups and that of the control (untreated) group were shown in Table 6. Syncytia formation could be detected when virus supernatant from HTLV-IIIB/H9 was incubated at room temperature for 1, 2 and 4 hours; and at 37OC for 30 and 60 minutes whereas incubation of the virus at 56OC for 30 and 60 minutes and at 65OC for 15, 30 and 60 minutes can eliminate the syncytia formation induced by HIV infection. Five and seven cycles of freeze and thaw treatment of the virus can reduce the syncytia formation on Day 3 to some extent when compared to the untreated samples, but it is still detectable.
The DNAs were purified from Day 3 cultures without syncytia formation and subjected to PCR analysis. After amplification by env primers, the products were analyzed on a 3% agarose gel. Fragments of predicted size (141 bps) were detected on the samples with incubation at 56OC for 30 and 60 minutes and at 65OC for 30 and 60 minutes, respectively (Figure 2). When another set of primers specific for gag region of HIV-1, the same results emerged (data not shown).
Given the medical and social significance of a positive test for HIV antibody, test results must be accurate and interpretations of the results must be correct. For these reasons, the Department of Health has requested that an individual be considered to have serological evidence of HIV infection only after an ELISA test is repeatedly reactive or a PA test is positive and another confirmatory test such as WB has been performed to validate the results.
Appropriate storing and shipping conditions of specimen are the key points of accurate testing results for detection of infections. For viral infections, the detection of corresponding antibody responses is easy and reliable [8]. However, because the major composition of an antibody is amino acid which is heat liable, high temperatures may cause denaturation of proteins and results in false negative outcome. Therefore, it is important to ensure that the samples to be tested are at good quality and under good conditions.
According to the results here, high temperature up to 65OC had no obvious effect on the sensitivity of PA and WB in detection of anti-HIV-1 antibody. In ELISA, only the treatment of 65OC for 60 minutes did turn the result to negative by the Abbott ELISA kit but not by the other two ELISA kits. Shirazian et al. [9] found that the washed samples of residual blood from needles and broken pieces of glass that had been exposed to HIV-1-seropositive sample and left at room temperature for one hour, one day and one week still resulted in positive tests for HIV-1 antibody by ELISA, immunofluorescence (IF) and WB. Their results are consistent with the findings here; both indicated that HIV-1 antibody is quite stable in serum.
On the other hand, results here differed from the results of experiment performed by Evans et al. [10] and McBride et al. [11], those reports indicated that heat-inactivation of serum specimens before screening by ELISA for HIV antibody can give false positive results and can also interfere with WB analysis. The major reason for the difference may be the detection kits used in this study are much more advanced than the old generation kits.
An extra fifty normal serum samples were tested by ELISA (Wellcome and Abbott) and PA. All those testing results were still negative when the serum was pre-heated to 56OC for 30 minutes. New generation kits have improved significantly in quality, specificity and sensitivity over the old kits used several years ago. Therefore, the false positive problem caused by the heating process may have been resolved or improved already.
Currently, only WB is used as confirmatory test in Taiwan. Although all WB interpretations are based on detecting antibodies against specific viral proteins, different official agencies or institutions have promoted the use of different sets of criteria for interpreting HIV-1 band patterns in the WB test. So far, CDC and World Health Organization (WHO) interpretive criteria [12,13] are those most accepted by the testing laboratory. Both WB interpretive criteria consider a WB test that has no bands as nonreactive for HIV antibody, and WB band patterns that do not meet the specific criteria for reactive are termed " indeterminate". For CDC criteria, any two bands of p24, gp41 or gp120/160 are considered as reactive; for WHO criteria, any two of the env proteins (gp41, gp120 and gp160) are regarded as positive. Therefore, p24, gp41, gp120 and gp160 should deserve more concern during the processing of samples, because if one of these bands has disappeared, the result may become "indeterminate". Data here suggested that the experimental treatments did not lessen the intensity of the bands. The serum samples used for the HIV-1 antibody proficiency test were delivered by this staff and kept carefully in container packed with ice. Even the worst case, they would not have been left at room temperature for more than two to three hours. Therefore, participating laboratories doing proficiency tests of HIV-infection should not complain again that poor storage conditions have influenced their testing results.
For purposes of public health, people usually pay much more attention to false negative results for HIV infection. Factors that were or might have been responsible for the appearance of false negatives are (1) assay insensitivity; (2) window period [14] and a late stage of AIDS with no antibody response yet, and antibodies saturated by an overburden of HIV antigens, respectively; (3) incorrect original designation of the specimen as positive or continuing uncertain serostatus; (4) deterioration of the specimen in storage, and (5) technical error. To avoid false negative results, the Department of Health has issued different policies or strategies such as strict regulations on detection kits before their marketing, regular check of the detection kits already on the market, proficiency test for detection of HIV-infection, training programs for laboratory technicians and others. Even through all the efforts have been made, the problem of a window period of HIV infection can not be overcome. When whole-virus-lysate ELISA (old generation kit) was used to screen blood donations from 1985 to 1990 in the United States, the average length of the window period was 45 days [14]. The average window period of the new generation kit, which can detect both IgG and IgM HIV-1 and HIV-2 antibodies, is 25 days [15,16]. During the acute period of infection, tests for p24 antigen can detect HIV infection earlier than antibody tests [17]. p24 antigen, the core structural protein of HIV, is detectable two to three weeks after HIV infection. On an average, p24 antigen is detected an estimated six days before antibody tests become positive [18,19]. However, the detection kits can not be depended upon totally to resolve problems caused by a window period. Even through the sensitivity of detection kits has been improved significantly, the answer to window period still remains unsolved. An improvement in donor interviewing about the behaviors associated with risk for HIV infection is one of the best ways to further secure safety of blood bank and public healths.
Since temperature has no effect on PA, ELISA and WB sensitivity in detecting HIV-1 antibody, an attempt was made to heat the serum samples before the process of detection to inactivate the virus and reduce potential risk of exposure to HIV-infection. Both syncytia formation assay and PCR were applied to detect HIV infection. The HIV-infected cells can attract uninfected T cells bearing CD4 nearby to form a giant cell, a so-called "syncytia", which is very easy to observe under the microscope. PCR is an extremely sensitive method to pick up any signal (DNA or RNA) existing in samples. In this study, no sign of HIV-infection was observed by syncytia formation assay after a three-day incubation. Usually HIV-infected culture in vitro can induce CPE in two to four days when co-cultured with uninfected CD4+ T cells. Further analysis of the samples by PCR indicated that the HIV-1 signal can be detected. However, if the virus loads are extremely low, it may take longer for syncytia to form. Thus, the testing samples may not have been incubated long enough to see the CPE induced by HIV, but its genomic signals could still be picked up by PCR. High temperature can cause dissociation of the non-covalent bonds between gp41 and gp120, which in turn will reduce the syncytia formation ability of HIV. In addition, it is possible that syncytium-inducing (SI) property of HIV changes to non-SI by treatment with heat. All of these factors might contribute to the opposite results seen in CPE and PCR.
The report of Stromberg et al. [20] indicated that mild heat treatment (45OC) is not an adequate process for viral inactivation of red cell products. Putting the data together, results here suggested that HIV infectivity can be reduced to some extent by heating to 56OC for 30 minutes and 60 minutes or 65OC for 30 and 60 minutes, but it is very difficult to totally destory HIV-infectivity by such treatments.
The virus titer was not determined in this study; however, it should be higher in the supernatant collected from HTLV-IIIB/H9 cell culture than in the serum of HIV-positive individuals. Thus, it is reasonable to predict that after heat treatments of samples at 56OC or 65OC, the virus titer in any suspected serum should drop to even lower levels.
Only fifty samples were tested by two different ELISA kits with no apparent dissimilitude, therefore, the possibility of false positive results caused by heating of serum samples could not be excluded. Based on these findings, it is suggested that serum samples could be heated first under some circumstances such as to achieve an efficient proficient testing program to inactivate the virus or to reduce its infectivity before the process of detection of the HIV antibody. But laboratory workers are still urged to follow the universal precautions, with the basic assumption that all blood should to be considered to be potentially infective [21,22].
Acknowledgements
We thank Dr. Chin Li-Te at the Chinese Blood Services Foundation and Mr. Yeh Jane Shiann at Provincial Tao Yung Hospital for performance of the ELISA analyses.
|
Copyright: 1997, Chinese Medical Association (Taipei)