Proliferation and telomere length in acutely mobilized blood mononuclear cells in HIV infected patients

Abstract

The aim of the study was to investigate the mobilization of T cells in response to a stressful challenge (adrenalin stimulation), and to access T cells resided in the peripheral lymphoid organs in HIV infected patients. Seventeen patients and eight HIV seronegative controls received an adrenalin infusion for 1 h. Blood was sampled before, during and 1 h after adrenalin infusion. Proliferation and mean telomere restriction fragment length (telomeres) of blood mononuclear cells (BMNC) and purified CD8⁺ and CD4⁺ cells were investigated at all time points. In patients, the proliferation to pokeweed mitogens (PWM) was lower and decreased more during adrenalin infusion. After adrenalin infusion the proliferation to PWM was restored only in the controls. In all subjects telomeres in CD4⁺ cells declined during adrenalin infusion. Additionally, the patients had shortened telomeres in their CD8⁺ cells, and particularly HAART treated patients had shortened telomeres in all cell-subtypes. The finding that patients mobilized cells with an impaired proliferation to PWM during and after adrenalin infusion has possible clinical relevance for HIV infected patients during pathological stressful conditions, such as sepsis, surgery and burns. However, this study did not find a correlation between impaired proliferation and telomeres. It is concluded that physiological stress further aggravates the HIV-induced immune deficiency.

INTRODUCTION

Infection with human immunodeficiency virus (HIV) causes both quantitative and qualitative T cell defects in the host. The quantitative defect includes mainly a numerical affection of cells within both central and peripheral lymphoid organs and the peripheral blood [1,2]. The qualitative defect comprises a significant restriction in the CD4⁺ T cell repertoire [3–5], impaired proliferative responses to mitogens and antigens [6–8] and a significant decline in telomeres in primarily CD8⁺ cells and later possibly also in CD4⁺ cells [9–15]. The central immunological disturbances in HIV infection make it attractive to investigate immunological markers that best uncover both the basic pathological mechanisms and the clinical consequences. Investigation of telomeres that are unique protein-DNA complexes at the endings of linear chromosomes has been suggested as a tool to describe lymphocyte replicative history. In humans, the DNA component of telomeres consists of hexanucleotide repeats (TTAGGG)_n of approximately 5–15 kilo bases (kb) [16,17]. Their function appears to include protection of chromosomes from illegitimate fusion, the localization of chromosomes in the nucleus and selective silencing of subtelomeric genes [18]. Due to the semiconservative DNA replication, approximately 50–200 base pair (bp) at the terminal of the lagging strand is left unreplicated with each cell division. In the absence of compensatory mechanisms to counter for this effect, namely the telomerase enzyme, a continued shortening of the chromosomes (telomeres) thus occurs with an average loss of 33 bases each year in leucocytes [19]. Telomeres thus decrease in length with both age and proliferation in somatic cells [17,20,1,2,3, providing a measurement not only of the prior replicative history but possibly also of the potential for further cell division [20,22].

The majority of clinical experimental studies investigate immunological parameters in the peripheral blood because this is most easily accessed. However, investigations of peripheral blood have some limitations since only 1–2% of the entire T cell pool is present [23,24].

Previously we have used adrenalin infusion as an immunological tool to investigate both the acute mobilization of immune competent cells in response to a stressful challenge and to characterize immune competent cells resided in the peripheral lymphoid organs in HIV infected patients [25,26]. The advantage of this approach is the opportunity to investigate a functional immunological response to a situation partly mimicking real-life stressful conditions such as sepsis, surgery and burns [27]; furthermore, to access more T cells than the 1–2% in the peripheral circulation [28]. For instance, a threefold increase in the lymphocyte count observed in our experiment [26] enables us to investigate approximately 4–8% of the entire T cell pool.

The aim of the present study was to investigate the proliferative capacity and the telomere lengths in acutely mobilized immune competent cells in HIV infected patients. We hypothesized that the proliferation in HIV infected patients would be further aggravated by a stressful challenge. Additionally, we wanted to investigate if a possible aggravation in proliferation was explained by the mobilization of cells with shorter telomeres. Due to the great impact of highly active antiretroviral therapy (HAART) on HIV infection [29] both HAART treated and untreated patients were enrolled in this study, together with a healthy control group.

MATERIALS AND METHODS

Study subjects

The study subjects and the experimental protocol have been described previously [25,26]. Briefly, 17 HIV infected patients and eight HIV seronegative controls gave written informed consent to participate in the study and receive adrenalin infusion for 1 h. The HIV infected patients included nine patients on stable highly active antiretroviral therapy and eight untreated patients, chosen so these two groups had comparable CD4 counts (Table 1). Control subjects were age- and gender-matched to the patients (Table 1).

Table 1.

Data from the HIV infected patients (individual or mean (s.d.)) and the control group (mean (s.d.))

Group and subject i.d.1		n	Gender	Age [years]	First test [years]	CD4 counts3	HIV RNA6	Nadir CD4 counts	Max HIV RNA	HAART regimen5	Months HAART6	Previous treatment7
HIV + HAART		9	9 M	44	8·6	3733a	1·69	238	4·63	–	–	–
(s.d.)				(8)	(4·5)	(107)	(2·33)	(67)	(0·55)
01			M	41	1	275	1·85	95	5·24	AZT, 3TC, IND	10
02			M	56	13	256	< DL	262	< DL	3TC, d4T, IND	6	AZT, ddI
03			M	31	1	347	2·02	252	4·89	AZT, 3TC, RTV, SAQV	8
07			M	45	11	449	< DL	294	5·33	AZT, 3TC, IND	8
08	T-082		M	46	11	398	1·53	255	4·93	3TC, d4T, RTV, SAQV	6	1 year AZT
10	T-10		M	40	12	574	< DL	261	4·21	3TC, d4T, RTV	9	9 years AZT
11			M	55	9	419	3·32	316	4·38	AZT, RTV, 3TC	8
12			M	44	11	238	< DL	172	4·32	3TC, d4T, RTV	7	AZT
15	T-15		M	37	8	402	< DL	233	3·77	AZT, 3TC, RTV	11	2 years ddI
HIV + untreated		8	7M, 1F	37	7·6	3703b	4·39	312	4·34	–	–	–
(s.d.)				(7)	4·9)	(189)	1·83)	(141)	0·71)
04			M	41	15	225	4·60	158	4·72
06			F	47	5	539	4·72	119	4·37
09			M	27	3	174	5·39	247	5·31
13	T-13		M	39	<1	450	4·65	260	3·99
14	T-14		M	34	6	575	4·08	525	4·28
17			M	28	7	500	4·19	444	4·06
23			M	50	13	123	4·70	340	4·97
24	T-24		M	40	11	399	2·75	399	3·00
Controls8		8	7M, 1F	37	–	656	neg	–	–	–	–	–
(s.d.)					(7)	(366)

M = male, F=female, DL= detection limit, NT=not treated, nadir = lowest CD4 count measured, max = maximum HIV RNA measured, neg = negative HIV test, 3TC = lamivudine, d4T = stavudine, RTV = ritonavir, SAQV = saquinavir, AZT = zidovudine, ddI= didanosine, IND = indinavir.

¹Telomere lengths (T) were measured in subjects no: 08, 10, 15, 13, 14, 24, 16, 18, 20, 21.

²i.d. Legends used in the figures.

³CD4⁺cell counts at t = 0 are shown cells ×ul^–1. Mean (s.d.) CD4⁺ cell counts 12, 6 and 3 months before the trial, in the HAART treated patients

^3a293(125), 333(69) and 412(153), respectively, ] and the untreated patients

^3b461(59), 374(85) and 461(90), respectively, ].

⁴HIV RNA in plasma at t = 0 [Log₁₀ copies × ml^–1 ].

⁵Present HAART regimen and months on this regimen at t = 0

⁶(mean 8·1 month (s.d. 1·7)).

⁷Previous antiretroviral treatment before attending to HAART.

⁸Ages in the four HIV sero-negative controls in whom telomere lengths were measured: subject i.d. T-16 (28 years), T-18 (42 years), T-20 (43 years), T-21 (44 years).

Eperimental protocol, collection of samples and cell preparation

Adrenalin (adrenalin, 1 mg ×ml^–1, DAK, Denmark) was diluted in isotonic saline and infused at a rate of 100 ng ×kg^–1×min^–1 for 1h. Blood and plasma were sampled at the baseline, during adrenalin infusion (after 15 and 45 min, maximum leukocytosis), and 1 h after adrenalin infusion had stopped (maximum leukopenia). Blood was sampled from an indwelling venous catheter and collected in glass tubes containing heparin (non-conserved 5000 IE × ml^–1, SAD, Denmark). Blood mononuclear cells (BMNC) were isolated by density gradient centrifugation (Lymphoprep Nycomed Pharma AS, Oslo, Norway) on LeucoSep tubes (Greiner, Frickenhausen, Germany) and washed three times with Stock solution I and HANK (Bie & Berntsen AS, Rødovre, Denmark). The BMNC were frozen in freezing medium and kept in liquid nitrogen until thawed for analysis. Thawed cells were used subsequently in the assays because previous studies had shown that thawing had a limited effect on the proliferative response and that the day-to-day variation in the proliferation assays was reduced by thawing cells and performing the analyses in patients and controls in parallel. The experimental protocol was approved by the local ethical committee for Copenhagen and Frederiksberg Communities (project no. KF 02–050/97).

Proliferation assays

After thawing, cell cultures were performed in triplicate in 96-well microtitre plates (Nunc, Brand Products, Roskilde, Denmark), in which 59·4×10³ cells were resuspended in 200 μl per well and incubated for either 72 h or 192 h. The 72 h incubation examined the radioactive thymidine uptake in cell cultures stimulated with either (a) medium alone, (b) pokeweed mitogens (PWM) (Gibco BRL Life Technologies, Gaithersbourg, USA) 2×10³ times diluted stock solution or (c) phytohaemaglutinin (PHA) 20 mg ×ml^–1 (Difco Laboratories, Detroit, Michigan, USA) and (d) interleukin-2 (IL-2) 20U/ml (Boehringer Mannheim, Germany). The 192-h incubation examined the radioactive thymidine uptake in cell cultures stimulated with (a) medium alone or (b) ethanol inactivated Candida antigen (an optimal concentration was calibrated in our laboratory). During the last 24 h, cells were exposed to ³H ]thymidine. The cultures were collected on glass fibre filters with a micromate 196 harvesting machine (Skatron, Lierbyen, Norway) and the ³H ]thymidine incorporation was measured in a liquid scintillation counter (TRI-CARB Packard Instruments Rockville, MD, USA). For each triplicate, lymphocyte proliferation was recorded as the mean counts per minute (cpm). Stimulation Indexes (cpm_stimulated/cpm_NaCl) are presented as geometric mean with 95% confidence interval.

Isolation of CD4⁺ and CD8⁺ T cells

After thawing, BMNC were resuspended in phosphate buffered saline (PBS, Bie & Berntsen AS, Rødovre, Denmark) and separated in CD4⁺ and CD8⁺ T cells by positive selection with Dynabeads M-450 CD4 (clone ST4) and M-450 CD8 (clone ITI-5C2) (Dynal, Oslo, Norway). The bead:cell ratio was 3:1, and each subtype of Dynabeads was in a final concentration of at least 2×10⁷ Dynabeads ×ml^–1. The calculated number of beads were prewashed with PBS buffer and resuspended in the cell suspension with gentle tilt incubation for 45 min at 4°C. The CD4⁺ or CD8⁺ cells were finally trapped by a magnet (Dynal MPC-6, Dynal, Oslo, Norway) and washed three times with PBS buffer.

The purity of the magnetic cell isolation was investigated by flow cytometric analyses on ungated BMNC, CD4⁺ or CD8⁺ T cells using monoclonal antibodies (MoAbs) against CD3, CD4 and CD8 antigens in a triple-colour mixture. Direct cell surface immunofluorescent staining was performed on 10⁵ BMNC, CD4⁺ or CD8⁺ T cells according to the manufacturer’s recommendations. The MoAbs used were directly conjugated to fluorescein isothiocyanate (FITC), phycoerythrin (PE) or R-phycoerythrin-cyanin-5·1 (PeCy5). The MoAbs were obtained as follows: IgG₁ isotype control (FITC-PE-PeCy5, Immunotech, Inc., Westbrook, Maine, USA), CD3 (DAKO, Glostrup, Denmark, clone UCHT1), CD4 (DAKO, clone MT310) and CD8 (DAKO, clone DK25). The subsequent flowcytometric analyses were performed using an Epics XL-MCL flow cytometer (Coulter, Florida, USA). The magnetic cell separation yielded a purity of 97·2% and 97·8% within the CD4⁺ and CD8⁺ cell subset, respectively (data not shown).

DNA purification and determination of telomere restriction fragment (TRF) length

DNA was purified from BMNC, CD4⁺ and CD8⁺ T cells with a Quiagen blood kit (Hilden, Germany). Briefly, isolated non-detached CD4⁺ and CD8⁺ T cells together with BMNC were all resuspended in 200 μl PBS buffer. A volume of 25 μl Qiagen protease and 200 μl AL buffer were added and incubated 10 min at 70°C. Hereafter 210 μl ethanol (96–100%) was added and Dynabeads were removed from the CD4⁺ and CD8⁺ T cell solution by placing the test tube in a magnet (Dynal MPC-6) for 5 min, before transferring the DNA from the bottom of the tube to a QIAamp spin column. After centrifugation at 6000g for 1 min, the collection tube was discarded, and the QIAamp spin column was washed twice with 500 μl AW buffer (centrifuged first at 6000g for 1 min followed by full speed (20 800 g for 3 min). After replacing the QIAamp spin column in a microfuge tube, the DNA was eluted with 200 μl AE buffer preheated to 70°C and incubated for 5 min at 70°C, before being centrifuged at 6000 g for 1 min. DNA concentration was measured by UV absorbans before it was stored at – 80°C until analysed.

A total of 200 ng thawed DNA was digested with HinfI (10 U) (Gibco BRL) and ethanol precipitated. The samples and two DNA size markers, 1 kb and 5 kb (Gibco BRL) were loaded onto a 0·5% agarose gel and separated by electrophoresis. The DNA was blotted to a positive nylon membrane (Appligene Oncor, Illkirch, France), and subjected to hybridization with a ³²P-labelled (TTAGGG)₄ telomeric probe (Pharmacia, Biotech, Uppsala, Sweden) and ³²P-labelled 1 kb and 5 kb markers in 1 mm EDTA, 500 mm Na ] HPO₄, 7% SDS. Hybridization was carried out overnight at 55°C. After two washes in 6 × SSC, 0·1% SDS at 55°C the membrane was exposed to a Phosphor Imager Screen (BIO RAD, California, USA). The mean TRF (mTRF) length was defined as ∑(Z_i× l_i)/∑(Z_i) where Z_i is the integrated signal in the interval i and L_i is the telomere length at the midpoint interval [19]. Since the results of the assay depended on a complete digestion of the non-telomeric DNA, 1 mg of λ DNA was added to each sample as a control. For complete digestion bands of particular size from the λ DNA were expected (visualized after EtBr staining on a UV-table).

Assay variability

To minimize assay variability, equal amounts of DNA were used and samples from the same subject were run on the same gel. The intra-assay and interassay coefficient of variation of the assay were 1·1% and 3·0%, respectively.

Plasma HIV RNA

Plasma HIV RNA was determined with Roche Amplicor HIV Monitor test (Roche Diagnostics, Nutley, NJ, USA). The lower detection limit was 20 copies ×ml^–1. Data are presented as log₁₀ copies HIV RNA ×ml^–1.

Statistics

Data were tested for the effect of ‘time’ and ‘HIV-serostatus’ (HIV/controls) in a two-way analysis of variance (anova). Data from the patients were further tested for the effect of ‘time’ and ‘HIV treatment-status’ (⁺HAART/÷HAART) in a similar two-way anova. The model was:

The residuals in the anova were examined for normal distribution through investigation of a histogram and a normal plot. If the residuals were considered not to be normally distributed, data were log transformed and investigated again. This was the case for the proliferation data. If the group/HIV-serostatus–time interaction was significant, the Dunn–Sidak-adjusted two-sample t-test was performed. If the effect of time or time and group/HIV-serostatus were significant, the Dunn–Sidak-adjusted paired t-test was performed.

To test if telomeres in different cell subpopulations differed in each group of subjects, the telomeres at t =0 for BMNC/CD8⁺/CD4⁺ cells were compared with paired t-tests. To test for a correlation between telomeres and proliferation or immunological characteristics, Pearson’s correlation with Dunn–Sidak adjustment was performed.

Data are presented as mean (s.d.) unless otherwise stated. Correlation data are presented with a P-value and R. The significance level in all tests were P < 0·05; however, P < 0·10 are also shown. Statistical calculations were performed using systat 8·0 for windows (SSPS Inc., Chicago, USA).

RESULTS

Immunological and virological characteristics of the HIV infected patients

No significant differences between the HAART treated and untreated patients with regard to their ages, CD4 counts, nadir CD4 counts, maximum viral load and years from first positive HIV test were observed (Table 1). However, a tendency towards higher ages and lower nadir CD4 counts were observed in the HAART treated patients (Table 1). As expected, the untreated patients had the highest viral load [26]. The subjects in the telomer groups were comparable.

Proliferation

In cell cultures stimulated with the mitogens PWM or PHA, the proliferation changed in response to adrenalin infusion in all subjects (P < 0·01 and P < 0·001, respectively) (Fig. 1).

Fig. 1.

Proliferation in cell cultures stimulated with (a) PWM, (b) PHA, (c) IL-2 or (d) Candida antigen estimated by the incorporation of radio labelled 3H ]thymidine. Stimulation indexes (cpm_stimulated/cpm_NaCl) are shown as geometric mean with 95% confidence interval. Each of the three subgroups is shown. *, ** and *** indicate difference from prevalue (two-way anova effect of ‘time’) with P < 0·05, P < 0·01 and P < 0·001, respectively. A suffix ^C or ^H indicates that only controls or HIV infected patients, respectively, differed from the prevalue in a paired t-test. † indicates difference between controls and HIV infected patients with P < 0·05 in both anova effect of ‘group’ and Dunn–Sidak-adjusted two-sample t-test. ■, HIV + HAART treated; ▴, HIV + untreated; •, controls.

In the HIV infected patients the cell cultures stimulated with PWM had a lower proliferation compared with the controls (P < 0·05). After 15 min adrenalin infusion the proliferation to PWM declined (P < 0·01) and remained low in the patients. After the adrenalin infusion the controls achieved a marked increase in proliferation exceeding baseline values (P < 0·05). In contrast, the patients did not achieve a similar increase in PWM stimulated proliferation and the pre-existing difference in PWM proliferation was thus further aggravated in response to adrenalin infusion. The PHA responses appeared visually similar to the PWM responses with a decline in the proliferation during the adrenalin infusion, although no significant differences between patients and controls were found. In cell cultures stimulated with Candida antigen, neither differences between groups nor responses to adrenalin infusion were observed. The finding that neither PHA nor Candida stimulated cell cultures differed between patients and controls is surprising, but explained most probably by a large physiological variation in these assays and a small sample size.

In contrast to the cell cultures stimulated with mitogens, proliferation in the IL-2 stimulated cultures increased in all subjects during and after adrenalin infusion (P < 0·001) without differences between patients and controls. The adrenalin-induced increase in proliferation in IL-2 stimulated cell cultures is explained most probably by an adrenalin-induced mobilization of IL-2 responsive CD3⁻ natural killer cells. Consequently, the decline in proliferation in PWM and PHA stimulated cell cultures is explained most probably by a concurrent decline in the proportion of CD3⁺ cells [25,30].

Telomere lengths in BMNC, CD8 and CD4⁺ cells

The telomere lengths were analysed in only 10 subjects due to a limited amount of cells available. This should be considered when interpreting the results.

In the 10 subjects the mean telomere restriction fragment length (mTRF) declined in the CD4⁺ cells during adrenalin infusion (P < 0·01) (Fig. 2), whereas no significant decline in mTRF was observed in the CD8⁺ cells (P = 0·10). The mTRF in BMNC changed neither in patients nor controls (Fig. 2). When the patients and controls were compared these groups had comparable mTRF. However, in the HIV group the three HAART treated subjects had shortened mTRF in both BMNC, CD8⁺ and CD4⁺ cells (P < 0·05, P < 0·05 and P < 0·05, respectively) (Fig. 2). This finding is remarkable but should be taken with some caution due to the small sample size.

Fig. 2.

Telomere restriction fragments length in (a) BMNC, (b) CD8⁺ and (c) CD4⁺ cells. The mean values (without error bars for clarity) at the left and data for each of the 10 subjects at the right. Mean mTRF (s.d.) at t = 0, t = 15, t = 45 and t = 120 in controls (BMNC: 8·3(1·2), 8·2(1·1), 8·0(0·9), 7·9(0·8); CD4⁺: 8·4(1·1), 8·1(1·1), 7·7(1·6), 7·9(1·7); CD8⁺: 8·1(0·7), 7·6(1·3), 7·6(1·1), 7·5(0·6)), HAART treated (BMNC:7·9(0·8), 7·1(0·5), 6·9(0·6), 7·2(0·2); CD4⁺: 7·3(0·3), 7·2(0·2), 6·9(0·4), 7·0(0·4); CD8⁺: 6·7(0·5), 6·1(0·4), 6·3(0·7), 6·5(0·7)) and untreated (BMNC: 8·5(0·2), 8·1(0·2), 8·3(0·7), 8·4(1·3); CD4⁺: 8·7(0·8), 8·7(1·0), 8·4(0·8), 7·9(0·7); CD8⁺: 8·0(0·8), 7·8(0·5), 7·8(0·4), 7·9(0·8)) patients, respectively. ¶ indicates difference between HAART treated and untreated HIV infected patients with P < 0·05 anova effect of ‘group’. ––, HIV + HAART; - - -, HIV untreated; —, controls. –•–, T-08; –▴–, T-15; -■-, T-14; —•—, T-16; —▴—, T-20; –■–, T-10; -•-, T-13; -▴-, T-24; —■—, T-18; —— T-21.

To test if the mTRF influenced the proliferation, we tested for a linear correlation between mTRF (BMNC, CD8⁺, CD4⁺) and proliferation (PWM, PHA, IL-2, Candida antigen) at baseline (t = 0), during (t = 45) and after (t = 120) adrenalin infusion. The only significant correlation, after Dunn–Sidak adjustment, was a positive correlation between proliferation to IL-2 at baseline and mTRF in CD4⁺ cells (R = 0·84, P < 0·05) and a similar tendency in BMNC (R = 0·81, P = 0·09) (data not shown). Thus, we found no evidence for a connection between impaired/reduced proliferation and shortened mTRF in this study. In the patients we tested for a linear correlation between baseline mTRF (BMNC, CD8⁺, CD4⁺) and nadir CD4 count, maximum viral load, years from or age at first positive HIV test. We found a strong positive correlation between nadir CD4 counts and mTRF in CD4⁺ cells (R =0·950, P < 0·01) and a similar tendency in CD8⁺ cells (R = 0·79, P = 0·06) (Fig. 3). Finally, we investigated if the mTRF at baseline differed between subtypes of cells in patients and controls. As described previously, the patients had shorter mTRF in their CD8⁺ cells compared with both CD4⁺ cells (P < 0·05) and BMNC (P = 0·07) whereas no such differences were found in the controls.

Fig. 3.

Correlation between nadir CD4 counts cells ×ul^–1 ] and telomere length in CD4⁺ (R = 0·95, P < 0·01) and CD8⁺ cells (R = 0·79, P = 0·06) in six HIV infected patients (i.d.: T-08, T-10, T-15, T-13, T-14, T- 24). ○, CD4 mTRF t = 0; ×, CD8 mTRF t = 0.

DISCUSSION

In the present study we used adrenalin infusion to investigate immune competent cells mobilized from peripheral lymphoid organs in response to a stressful challenge. The main finding was that adrenalin infusion aggravated the pre-existing defect in PWM proliferation in the HIV infected patients.

The finding that HIV infected patients mobilized cells with an impaired proliferation to PWM is interesting, and suggests that the HIV-induced immune defect is further aggravated during conditions with an increased adrenalin level. This could be of possible clinical relevance in HIV infected patients during pathological stressful conditions, such as sepsis, surgery and burn [27] in accordance with previous findings [31]. In a prior study we showed that adrenalin preferentially mobilizes T cells with an activated/memory phenotype and, furthermore, that the adrenalin-induced mobilization of activated/memory cells mirrors the proportion of these cells seen in a resting blood sample in both HIV infected patients and healthy controls [26]. The impaired PWM response observed in the HIV infected patients could be explained by the mobilization of many preactivated cells with a lower proliferative capacity possibly caused by a longer replicative history (and shorter telomeres). However, the only correlation found between proliferation and telomeres was between telomeres in CD4⁺ cells and IL-2 stimulated proliferation. Consequently, it is most probable that the impaired PWM response in HIV infected patients mirrors a HIV-induced immune defect not caused by a shortening in telomeres [6]. However, it is possible that the previously described reduced IL-2 stimulated proliferation in HIV infected patients [7,8] could, at least partly, be caused by the presence of CD4⁺ cells with shortened telomeres.

Remarkably, the present study found neither differences in the PHA and Candida-induced proliferation comparing patients and controls nor differences in the proliferation comparing HAART treated and untreated patients. The latter finding is in contrast with some previous findings from our group [32[. However, the lack of immunological restoration in the HAART treated patients in the present study could be caused by a short period of antiviral treatment (Table 1) or immunological/virological differences between the HAART and untreated HIV infected patients (see discussion below). Finally, the normal physiological variation in a stress-response possibly explains much of the large variation in the proliferation assays, and could hide some smaller differences between the three groups.

Besides the small sample size, this study found that the three HAART treated patients had shortened telomers in several cell-subpopulations compared with the three untreated HIV infected patients. This finding is in accordance with some previous studies [15,33], but in contrast with others [34]. However, due to the sample size these results should be interpreted with caution. In the present study the telomere length analyses from each subject was analyzed on the same gel which is why these data best describe intrasubject differences (between subpopulations of cells or between different time points). Low inter-assay coefficient of variation (3·0%) in the assay supports the validity of the observed telomere differences. However, the shortened telomeres in the HAART treated patients could have several explanations.

One simple explanation is a coincidental biological variation in these few subjects, e.g. caused by variability in telomerase activity [35]. Moreover, HAART treated patients could have a more aggressive history of HIV infection causing the initiation HAART. Additionally, aggressive HIV disease progression itself causes pronounced perturbation in both thymic output (less naive T cells), peripheral T cell expansion (more activated/memory T cells) and turnover of peripheral T cells (preferentially dying of activated/memory T cells) [15,36]. Obviously, these immunological disturbances could all influence telomere lengths measures in bulk CD4⁺ or CD8⁺ cells primarily because of the unequal telomere lengths in memory and naive T cells. We found a correlation between nadir CD4 count and telomeres in CD4⁺ cells, and it is possible that the slightly lower nadir CD4 counts (although not significant) in the HAART treated patients could explain a part of the shortened telomeres in these patients. Another explanation could be the slightly higher ages at the trial or at seroconversion in the HAART treated patients (although not significant) leading to a shortening in telomeres, reduced thymic output and more cells with a longer replicative history [36–38]. Finally, previous studies all agree that nucleoside analogues (especially AZT) have the potential to inhibit the telomerase enzyme (reverse transcripase activity) [33,39,1,2,3], resulting in enhanced shortening of telomeres. However, these studies disagree about the effect of this telomerase inhibition on the telomeres both in vivo and in vitro[33,39,1,2,3]. In the present study, the two patients that were treated with AZT for the longest period of time (Table 1) had the shortest telomeres in all cell-subsets. In this study no data are available from the HAART treated patients before their initiation of therapy, therefore making it difficult to explain our findings. Future studies are therefore definitively needed.

We found that adrenalin infusion caused a 0·5–1 kb decline in the telomeres in the CD4⁺ cells, and to a lesser degree in the CD8⁺ cells. Although a 0·5–1 kb change is less than the overall difference between HAART treated and untreated HIV infected patients, it emphasizes that even a brief shift in the phenotypic composition of T cells affects telomeres. Furthermore, this finding supports the contention that mainly activated CD4⁺ and CD8⁺ T cells are mobilized in response to the applied stressful challenge [25,26,28,42]. Finally, and in agreement with previous studies [9,10,14,43], we found that the telomeres in the CD8⁺ cells were profoundly shortened in the patients. As suggested by many others [9,10,14,36,43], this probably reflects HIV-induced affection of the thymus and peripheral expansion of CD8⁺ cells.

Measurement of telomeres is not an obvious measure of T cell population kinetics because this method does not discriminate between peripheral T cell proliferation and thymic output of naive T cells or take selective killing of older cells into account [14,15,17,20]. The aim of the present study was not to evaluate T cell turnover, but to investigate the acute mobilization of immune competent cells in response to a stressful challenge and furthermore to investigate immune competent cells resided in the peripheral lymphoid organs in HIV infected patients. In conclusion, the HIV-induced impaired lymphocyte proliferation response was further aggravated by conditions accompanied with increased adrenalin levels. Although the CD4⁺ and CD8⁺ cells mobilized to the blood in response to adrenalin infusion had shortened telomeres, this could not explain the decreased proliferation.