Secondary Logo

Journal Logo


Primary macrophages from HIV-infected adults show dysregulated cytokine responses to Salmonella, but normal internalization and killing

Gordon, Melita Aa,d; Gordon, Stephen Ba,e; Musaya, Lisaa; Zijlstra, Eduard Eb; Molyneux, Malcolm Ea,e; Read, Robert Cc

Author Information
doi: 10.1097/QAD.0b013e3282f25107
  • Free



Nontyphoidal salmonellae (NTS) are the commonest bacterial pathogens isolated from blood cultures in sub-Saharan Africa where HIV is prevalent, and NTS bacteraemia is overwhelmingly associated with HIV among adults in this setting [1–4]. Patients are susceptible to severe invasive and recurrent NTS infections when the peripheral CD4 cell count falls < 200 cells/μl. Mortality is very high (33–73%) in all reported series [1,4–6], and 42% of survivors suffer bacteraemic recrudescence, despite antibiotic treatment [6], probably attributable to intracellular persistence in tissue macrophages.

Macrophages are important in innate defence against infections by salmonellae, and defects in macrophage phagocytosis have been described in HIV-infected individuals. Kedzierska et al.[7,8] described a phagocytic defect in monocyte-derived macrophages for toxoplasma, related to impaired Fcγ-R signalling caused by HIV [8]. Koziel et al.[9,10] used alveolar macrophages to demonstrate reduced phagocytosis [9] and oxidative burst [10] in response to Pneumocystis jiroveci, both defects being related to reduced mannose receptor (CD206) binding. Conversely, augmented microbicidal activity against Mycobacterium tuberculosis[11] and Cryptococcus neoformans[12] have been described using alveolar macrophages from subjects with early-stage HIV.

Macrophage activation is critical to phagocytic and other defence functions [13] and is modulated by the type-1 cytokine network involving interleukin (IL)-12, IL-10 and interferon-γ (IFNγ). IFNγ has an established role in early and late clearance of Salmonella typhimurium in animal and cellular models [14–17]. We have previously demonstrated, using human monocyte-derived macrophages, that recombinant IFNγ enhances internalization and nonoxidative killing; it also alters production of IL-10 and IL-12 following infection with S. typhimurium[18]. Janssen et al.[19] have also shown that recombinant IFNγ promotes intracellular killing of S. typhimurium by human monocyte-derived macrophages [19]. IFNγ primes macrophages for increased IL-12 and reduced IL-10 secretion, and IL-12 is important in late control and survival in murine salmonellosis [20,21]. Humans with defects in genes for IL-12 p40 and receptor experience serious NTS infections [22,23]. Conversely, neutralization of IL-10 in mice improves host resistance to infection [24].

We have hypothesized that the high rate of recurrence of NTS in HIV-infected adults could result from HIV-related defects in internalization or intracellular killing of S. typhimurium by tissue macrophages, or from dysregulated or ineffective T helper 1 cytokine responses of tissue macrophages. In both cases, any observed defects could be the direct result of HIV infection of the macrophage or a result of reduced enhancement of macrophage effector functions by IFNγ, as CD4 lymphocyte counts fall with disease progression.

Human tissue macrophages mature from circulating monocytes in peripheral organ compartments and form heterogeneous resident populations that are difficult to isolate in human subjects. Alveolar macrophages are, however, uniquely accessible (by bronchoalveolar lavage) in sufficient numbers for ex-vivo experimental manipulation. Pulmonary involvement during invasive NTS disease in HIV occurs in up to a third of patients, caused by NTS or concurrent second pathogens [6], and NTS may be directly cultured from lung parenchymal tissue [25–27]. Although the lungs are not the portal of entry, they are one of several possible tissue sites commonly involved during invasive NTS disease where NTS may localize or persist. The present study, therefore, used ex-vivo primary human alveolar macrophages (huAM) from HIV-infected and uninfected adults to examine the effect of HIV upon internalization and intracellular killing of salmonellae, and cytokine responses of macrophages to salmonellae in vitro. Unprimed and IFNγ-primed huAM were used to examine if any macrophage defence defects observed could be reversed by IFNγ.

Materials and methods

Subjects and controls

Subjects were recruited by advertisement in Queen Elizabeth Central Hospital, Blantyre, from among hospital staff and visitors. Blood was drawn for HIV antibody testing (Unigold and Serocard; Trinity Biotech, Dublin, Eire), CD4 lymphocyte count (Trucount; Becton Dickinson, San Jose, California, USA) and plasma HIV viral load (Amplicor; Roche Molecular Systems, Branchburg, New Jersey, USA). Consenting Malawian HIV-infected and uninfected adults, without clinical evidence of intercurrent infections over several visits and who had a normal chest radiograph, were recruited for bronchoscopy and bronchoalveolar lavage [28].

This study was given ethical approval by the Research Ethics Committee of the Liverpool School of Tropical Medicine and Hygiene, UK, and by the College of Medicine Research Ethics Committee, Blantyre, Malawi. All subjects gave informed consent as approved by these committees.

Collection and priming of human alveolar macrophages

The huAM were separated from fresh bronchoalveolar lavage fluid on day 0 as described. Briefly, 200 ml of warm sterile saline was instilled into a right middle lobe subsegmental bronchus and fresh bronchoalveolar lavage fluid was collected into siliconized sterile glass on ice. The fluid was filtered through sterile gauze and centrifuged (102 × g at 4°C for 5 min). The supernatant was stored for HIV viral load measurement (Amplicor). The cellular pellet was resuspended at 1 × 106 cells/ml in RPMI 1640 with 10% heat-inactivated newborn calf serum (HI-NCS; Gibco BRL, Life Technologies, Grand Island, New York, USA) and antibiotics [penicillin (40 IU/ml), streptomycin (75 IU/ml), gentamicin (80 IU/ml)]. Cells were plated at 1 × 106 cells/well into 24-well trays and cultured at 37°C in 5% carbon dioxide for 24 h. Cells for fluorescent microscopy were prepared on sterile glass coverslips (BDH, Poole, UK). Cells were washed after 24 h (day 1), and the adherent population of huAM thereby selected was subsequently cultured in antibiotic-free RPMI with 10% HI-NCS. On days 1 and 3, huAM were primed with IFNγ for 72 and 24 h respectively, using 100 ng/ml (1000 IU/ml) human recombinant IFNγ (Pharmingen, San Diego, California, USA).

Preparation of bacteria and infection of alveolar macrophages

Salmoella enterica serovar Typhimurium (NCTC 12023) was cultured overnight (16 h) in standing low-osmolarity LB broth (Oxoid, Basinstoke, UK) at 37°C, opsonized in 10% heat-inactivated human AB serum for 30 min, then resuspended in RPMI with 10% HI-NCS. Triplicate wells of huAM that were primed, primed for 24 or primed for 72 h were infected or mock-infected on day 4 at a target multiplicity of infection of 50 (MOI50) (5 × 107 bacteria/well), confirmed by dilution plating. Wells were incubated without centrifugation for 30 min at 37°C in 5% carbon dioxide. The end of the 30 min period of capture/binding/internalization of bacteria by macrophages was denoted time as time 0. Coverslip-adherent huAM at this time were washed and fixed in 2% paraformaldehyde for fluorescent microscopy.

Fluorescent microscopy

To measure internalization of bacteria, fixed coverslips were sequentially washed and stained for 12 min with 1: 20 rabbit anti-S. typhimurium poly-A antibody (Difco, BD Biosciences, Franklin Lakes, New Jersey, USA), then 1: 20 fluorescein isothiocyanate (FITC)-conjugated goat antirabbit IgG (Sigma, UK), and then counterstained with 4′6′-diamidino-2-phenylindole (DAPI) (Molecular Probes, Seattle, Washington State, USA) for bacterial DNA and cell nuclei. Slides were blinded and counted by a single investigator using a DMRB fluorescent microscope (Leica, Wetzlar, Germany). Cell-associated bacteria were defined as bacteria that lay within the limits of the cell cytoplasm. Cell-associated bacteria were further defined either as externally bound or internalized. DAPI-counterstained bacteria that were internalized were distinguished from externally bound bacteria by the absence of a ring of FITC-stained anti-IgG on internalized bacteria.

At times 0, 2 and 24 h, experimental and control huAM were stained with trypan blue for cell viability and density counts.

Intracellular bacterial viability assay

The gentamicin protection assay was used to measure bacterial viability following internalization. At time 0, excess bacteria were removed by washing, then huAM were incubated in media containing 100 μg/ml gentamicin for 30 min to kill remaining extracellular bacteria. As a control for gentamicin efficacy, paraformaldehyde-fixed huAM wells were infected and treated with gentamicin in the same manner, and residual bacteria counted. The standard for extracellular killing by gentamicin was set at > 1 log colony-forming units (CFU) residual extracellular bacteria below the viable count of bacteria subsequently released by lysis of live infected cells in experimental wells. Gentamicin 100 μg/ml was the minimum required to achieve this standard. In pilot experiments, stepwise increasing concentrations of gentamicin (50–400 μg/ml) gave increasingly efficient killing of external bacteria but constant yield of internalized bacteria upon cell lysis, suggesting that extracellular gentamicin did not compromise intracellular bacterial viability.

After gentamicin treatment for 30 min, media containing 10 μg/ml gentamicin was used to suppress ongoing replication and internalization. At 2 and 24 h, huAM were washed and lysed with 1% saponin, and the yield of CFU determined using a dilution plating method.

Cytokine measurement

Preliminary work demonstrated that the optimal duration of IFNγ priming of huAM for a secretory cytokine response to S. typhimurium was 24 h. The concentrations of IL-1β, IL-6 and TNFα at 2 h after challenge, and of IL-10 and IL-12 at 24 h after challenge, were measured in triplicate conditioned media by Cytokine Bead Array assay (Becton Dickinson), using a BD Facscalibur.

CD206 expression

HuAM surface expression of CD206 was assessed on IFNγ primed and unprimed huAM, detached on ice at day 4, using a buffer containing 1 mmol/l ethylenediaminetetraacetic acid and 0.5% lidocaine. Cells were stained with 1: 200 phycoerythrin-conjugated mouse IgG1 anti-CD206 or isotype control (Becton Dickinson) and analysed using a BD Facscalibur.

Statistical analysis

Data were analysed using Stata 8 (Statacorp, College Station, Texas, USA). All parameters were tested for normality of distribution using the Shapiro–Wilks test. Univariate comparisons of nonpaired and paired data at baseline were made using t-tests and the Mann–Whitney rank sum test. Multivariate analysis of variance (ANOVA) was used to examine the significance of HIV status, priming conditions and potential confounders. Correlation of endpoints with viral load and CD4 cell count was assessed by scatter graph with polynomial regression lines of best fit.



Alveolar macrophages were harvested from nine HIV-negative adults (median age 28 years; range, 18–53) and 11 HIV-positive adults (median age 36 years; range, 26–47). Median CD4 cell count among HIV-positive subjects was 250 cells/μl (range, 92–600). Median serum viral load from HIV-positive subjects was 1.8 × 105 copies/ml (range, 6900–2.9 × 106) and median bronchoalveolar lavage fluid viral load from HIV-positive subjects was 615 copies/ml (range, 0–1.4 × 104). Two HIV-positive subjects had undetectable HIV viral load in bronchoalveolar lavage fluid. All wells were inoculated at mean MOI50 (range 42–58).

Internalization of Salmonella typhimurium by alveolar macrophages

The effects of HIV status and of IFNγ priming on internalization by huAM of opsonized S. typhimurium were measured by fluorescent microscopy. There were no differences at baseline between unprimed huAM from HIV-positive and HIV-negative subjects in percentage of cells with cell-associated bacteria (P = 0.88), cell-associated bacteria/cell (P = 0.89), percentage of cell-associated bacteria internalized (P = 0.711) or internalized bacteria/cell (P = 1.00) (Fig. 1a). The effect of priming huAM with IFNγ was to reduce the percentage of cells with cell-associated bacteria, the number of cell-associated bacteria/cell and the percentage of cell-associated bacteria that were internalized (data not shown). These effects combined to give an overall reduction in the number of internalized bacteria per cell in IFNγ-primed huAM from HIV-negative donors (P = 0.005; Fig. 1a). The effect of IFNγ was greatest with longer duration of priming (Fig. 1a) and was seen over a range of MOI from 25 to 200 (data not shown). By ANOVA, the effect of IFNγ on internalization of bacteria was highly significant in huAM from both HIV-positive and HIV-negative donors (P < 0.00001) and there was no contribution of HIV status to internalization under different IFNγ priming conditions (P = 0.61). Among huAM from HIV-positive subjects, there was no correlation between internalization and CD4 cell count or viral load in serum or bronchoalveolar lavage fluid. HuAM from HIV-positive subjects, therefore, showed no difference in internalization of, or invasion by, S. typhimurium at baseline compared with cells from HIV-negative subjects, and they had fully preserved responses to IFNγ, which reduced internalization of bacteria.

Fig. 1
Fig. 1:
The effect of priming with interferon-γ (IFNγ) on internalization of Salmonella typhimurium by human alveolar macrophages (huAM) from HIV-negative and HIV-positive adults. (a) The number of internalized bacteria per cell measured by fluorescent microscopy, for huAM mock-primed or primed with IFNγ for 24 or 72 h, and challenged with S. typhimurium at a multiplicity of infection of 50. The graph shows medians (lines), interquartile ranges (boxes) and 5–95% ranges (whiskers). Data are derived from triplicate experiments for each condition, from nine HIV-negative and 11 HIV-positive subjects. (b) The internalization of S. typhimurium by huAM at MOI50, shown as geometric means and 95% confidence intervals (error bars). Data derived from triplicate experiments for each condition on huAM from four HIV-negative subjects.

In view of the known HIV-related defects in Fc-γR and CD206-mediated phagocytosis, further experiments were conducted using huAM from four healthy HIV-negative donors. Bacteria were opsonized or mock-opsonized and then inoculated at matched MOI. Expression of CD206 on primed and unprimed huAM was measured, and unprimed and IFNγ-primed huAM were preincubated for 30 min with 100 μg/ml mannosylated albumin (Sigma, Poole, UK) prior to bacterial challenge, to block CD206-mediated uptake of bacteria. Internalization of opsonized and nonopsonized S. typhimurium by huAM did not differ under unprimed or IFNγ-primed conditions (Fig. 1b). While priming with IFNγ did, as expected, result in reduced surface expression of CD206 (unprimed huAM: 35.8% of cells, mean immunofluorescence 9.17; huAM primed for 72 h: 20.2% of cells, mean immunofluorescence 6.8; P = 0.016), preblocking of CD206 did not alter internalization of bacteria, under unprimed or primed conditions, compared with controls (Fig. 1b). Taken together, these data suggest that opsonic or CD206-related uptake is not required for optimal internalization of S. typhimurium by huAM, and that the observed effect of IFNγ is not mediated by reduced expression of these candidate receptors.

Intracellular killing of Salmonella typhimurium by alveolar macrophages

The gentamicin protection assay was used to assess the intracellular killing of S. typhimurium by primed and unprimed cells from HIV-positive and HIV-negative subjects (Fig. 2). Yields of viable CFU were not significantly different in unprimed huAM from HIV-positive and HIV-negative subjects at 2 h (P = 0.38) or at 24 h (P = 0.63). The effect of priming with IFNγ was to significantly reduce the yield of bacteria at 2 h from huAM from HIV-negative subjects, suggesting increased early intracellular killing of bacteria. The effect of priming on intracellular killing was maintained until 24 h, and the magnitude of killing was greatest with longer duration of priming up to 72 h. IFNγ-primed cells also demonstrate reduced internalization of bacteria, and ANOVA was used to see if this accounted for the reduced yield upon lysis. Both the number of internalized bacteria (P < 0.0001) and IFNγ priming (P < 0.0001) were independent factors associated with reduced yield in huAM from HIV-negative subjects. ANOVA including HIV-positive and HIV-negative subjects showed no contribution of HIV (P = 0.23) but confirmed independent contributions of internalization (P = 0.0039) and IFNγ priming (P = 0.0039). Cell density was not different under different experimental conditions. Cell viability was > 95% under all conditions. Cell density and viability were not confounders in multivariate ANOVA (P = 0.13 and P = 0.42, respectively). Among huAM from HIV-positive subjects, there was no correlation between intracellular killing and CD4 cell count or viral load from serum or bronchoalveolar lavage fluid.

Fig. 2
Fig. 2:
The effect of priming with interferon-γ (IFNγ) on the yield of viable intracellular bacteria from lysed human alveolar macrophages (huAM) at 2 and 24 h after challenge with Salmonella typhimurium. HuAM from HIV-negative and HIV-positive adults were mock-primed or primed with IFNγ for 24 or 72 h. The graph shows medians (lines), interquartile ranges (boxes) and 5–95% ranges (whiskers). Data are derived from triplicate experiments under each condition, from nine HIV-negative and 11 HIV-positive subjects.

HuAM from HIV-infected subjects, therefore, showed no difference in intracellular killing of S. typhimurium compared with huAM from HIV-negative subjects. They retained the capacity to be primed by IFNγ for early killing and late intracellular control of Salmonella sp.

Cytokine production by alveolar macrophages in response to Salmonella typhimurium

Cytokine release by huAM (IL-1β, IL-6 and TNFα at 2 h, and IL-10 and IL-12 at 24 h) is shown in Fig. 3. Secretion of all cytokines by unchallenged control cells was very low. There were no significant differences in cytokine production between unprimed and primed control cells, or between control cells from HIV-positive and HIV-negative subjects. When challenged with S. typhimurium, cells produced significantly higher levels of all cytokines compared with control cells.

Fig. 3
Fig. 3:
Secretion of cytokines by human alveolar macrophages from HIV-negative and HIV-positive adults, at rest and after challenge with Salmonella typhimurium, under unprimed and interferon-γ (IFNγ)-primed conditions. Cells were mock-primed or primed with IFNγ for 24 h, then mock-challenged or challenged with S. typhimurium, and secreted cytokines measured from conditioned media. Interleukin (IL)-1b, IL-6 and tumour necrosis factor (TNF)-α were measured at 2 h postchallenge, and IL-10 and IL-12 were measured at 24 h postchallenge. The graphs show medians (lines), interquartile ranges (boxes) and 5–95% ranges (whiskers). Data are derived from triplicate experiments for both sterile and challenged cells, under each priming condition, from nine HIV-negative and 10 HIV-positive subjects.

The contributions of HIV status and IFNγ priming to cytokine production in response to S. typhimurium were examined by univariate analysis, and then by multivariate ANOVA, which showed that huAM from HIV-positive subjects produced more TNFα (P = 0.002), IL-10 (P = 0.01) and IL-12 (P = 0.01) than huAM from HIV-negative subjects. The contribution of IFNγ priming to cytokine production was not significant for these three cytokines (TNFα, P = 0.3; IL-10, P = 0.4; IL-12, P = 0.4). In contrast, the early proinflammatory cytokines IL-1β and IL-6 showed increased production in response to IFNγ priming (IL-1β, P = 0.02; IL-6, P = 0.01), but no significant effect of HIV status was seen (IL-1β, P = 0.3; IL-6, P = 0.1).

Correlation of cytokine production with CD4 cell count

Because of the markedly increased production of TNFα, IL-10 and IL-12 shown by huAM from HIV-positive subjects in response to S. typhimurium, the correlation between cytokine production and markers of HIV disease stage (CD4 cell count and viral load in serum and bronchoalveolar lavage fluid) was assessed using scatter plots with polynomial regression lines of fit. There was an overall trend of cytokine production seen in relation to peripheral blood CD4 cell count for TNFα, IL-10 and IL-12 (Fig. 4). Secretion of these cytokines was progressively higher among subjects with CD4 cell counts ranging from 600 down to 250 cells/μl. This corresponds to earlier disease, when CD4 cell counts are falling but are not yet AIDS defining. Among subjects with advanced disease and CD4 cell counts < 250 cells/μl, there was a rapid falling off of all cytokine responses to S. typhimurium challenge. This pattern was similar for all three cytokines and was seen among both unprimed and primed cells. There was no correlation seen between serum or bronchoalveolar lavage fluid viral loads and production of any cytokine.

Fig. 4
Fig. 4:
The relationship of cytokine secretion by human alveolar macrophages from HIV-positive adults to peripheral absolute CD4 cell counts following challenge with Salmonella typhimurium. Data are shown for the unprimed state (left) and for cells primed with interferon-γ (IFNγ) for 24 h. Each plotted data point is derived from triplicate experiments on a single subject. The graphs show polynomial regression lines of best fit (dashed), with confidence intervals (solid).


Bacteraemia from NTS causes a significant burden of severe recurrent disease among HIV-infected adults in Africa. The results described here are the first data acquired from primary human tissue macrophages in relation to NTS, and they reveal that internalization and intracellular killing, with and without IFNγ priming, are not impaired in macrophages from HIV-positive subjects. There are, however, major dysregulations of TNFα, IL-10 and IL-12 in cells from HIV-positive subjects following challenge with S. typhimurium, which is not reversed by priming with IFNγ. Furthermore, this dysregulation of cytokine production has an apparent pattern of relationship to CD4 cell count, seen consistently for all three cytokines.

Internalization of S. typhimurium by huAM was normal in cells from HIV-positive subjects, and priming with IFNγ enhanced resistance of huAM from both HIV-positive and HIV-negative subjects to this invasion. Although impaired phagocytosis of Toxoplasma sp. and Pneumocystis jiroveci has been described in macrophages from HIV-infected subjects, these defects involved Fc-γR or CD206, the mannose receptor, and we have demonstrated that neither of these routes of cell entry was necessary for S. typhimurium internalization by huAM. It is unsurprising, therefore, that we did not note any HIV-related defect in internalization. Independent of its effect on internalization, IFNγ also primed huAM for enhanced early intracellular killing and late control of Salmonella, and these responses were intact in huAM from HIV-infected subjects. Other reports have suggested a defective oxidative burst in macrophages from HIV-infected subjects in response to different pathogens, but we have previously demonstrated that IFNγ could prime human monocyte-derived macrophages for early nonoxidative killing of S. typhimurium[18]. Again, it is unsurprising that we did not note impaired killing of S. typhimurium by huAM from HIV-positive subjects. Our data, therefore, do not suggest that the intracellular persistence of NTS in HIV is attributable to defects in internalization or intracellular killing by HIV-infected tissue macrophages.

HIV infection of macrophages promotes a proinflammatory phenotype [29], with increased expression of multiple proinflammatory cytokine and chemokine pathways [30]. In particular, IL-6 and TNFα are released in excess from HIV-infected macrophages under a variety of stimuli [31–33], and enhanced production of IL-10 has also been described in HIV [34,35]. In contrast, peripheral blood mononuclear cells from subjects with HIV produce reduced levels of IL-12 upon stimulation [36,37].

We have demonstrated that huAM from HIV-positive subjects have unaltered resting cytokine production but oversecrete TNFα, IL-10 and IL-12 upon challenge with S. typhimurium. This may result from an altered alveolar milieu in HIV. There is a relative alveolar CD8 lymphocytosis observed in HIV, with IFNγ-producing CD8 cytotoxic T cells in earlier disease being later replaced by CD8 suppressor cells, which do not produce IFNγ [38,39]. If, however, altered in-vivo microenvironmental exposure of huAM to IFNγ in HIV explained our finding of enhanced ex-vivo cytokine production, these cells would also have shown IFNγ-modified internalization and killing at baseline ex vivo, which they did not. Another possibility is stimulation of huAM by latent opportunistic infections not detected by prebronchoscopy screening, such as tuberculosis, Mycobacterium avium complex or Pneumocystis sp. If that were the case, however, the cells would also produce excess cytokines at baseline ex vivo, and this was not observed.

It, therefore, seems likely that the overproduction of cytokines after challenge of huAM with S. typhimurium reflects other changes in the alveolar milieu, or an intrinsic effect of HIV on the cell. HIV subcomponents have been shown to alter signalling by mitogen-activated protein kinase, causing overproduction of TNFα and IL-10 [40,41], but there are no reports of overproduction of IL-12 in the context of HIV. The excess of IL-12 is particularly interesting, given its established role in protection against salmonellosis. IL-10 counterbalances the IL-12 pathway, with autocrine control of IL-12 by IL-10 in huAM from healthy individuals [42], but both cytokines were overproduced in parallel by huAM from HIV-positive individuals, and the ratio of proinflammatory cytokines to IL-10 did not change (Fig. 4). High levels of IL-12 in HIV apparently fail to confer protection from salmonellosis, probably because of dysregulation of the overall cytokine milieu, including in particular IL-10.

We further found a trend to a curvilinear relationship between secretion of these cytokines and CD4 cell count. Although the number of observations is small, the data suggest that macrophage cytokine production becomes globally increased, probably reflecting tissue macrophage activation as early HIV infection becomes established. As CD4 cell counts fall to < 250 cells/μl, however, there is a trend for a decline in the high levels of cytokine response. This corresponds to patients at an advanced stage of disease, who are particularly susceptible to invasive NTS infections. This novel finding now needs to be confirmed in other models and in larger numbers. CD4 cells activate macrophages through production of IFNγ, but in our experiments simple priming with IFNγ did not reverse the decline in cytokine responses at low CD4 cell counts, and a more complex interaction between macrophages and CD4 lymphocytes may be required. This is consistent with work by Pie et al.[43], who showed that late control of S. typhimurium in a mouse model required CD4 T cells, but that this clearance was by an IFNγ-independent mechanism.

In summary, our data show that IFNγ-primed and unprimed internalization and intracellular killing of S. typhimurium by huAM are intact in HIV. In contrast, there is overproduction of TNFα, IL-10 and IL-12, which cannot be easily explained by changes in exposure to IFNγ in the alveolar milieu, nor corrected by altered IFNγ-priming conditions ex vivo. This dysregulation and overproduction of critical cytokines is probably caused by chronic activation of macrophages as a direct consequence of infection with the HIV virus, or by other changes in the alveolar milieu during HIV infection. Furthermore, the dyregulation shows a trend towards a curvilinear relationship to CD4 cell counts, not previously described. We suggest that these dysregulations of cytokines, known to be critical for host defence against salmonella, result in a permissive environment for late invasive salmonellosis.


The authors would like to thank the patients and staff of Queen Elizabeth Central Hospital, Blantyre, Malawi.

Sponsorship: This work was funded by the Wellcome Trust (UK) with a Training Fellowship in Clinical Tropical Medicine (MAG), a Career Development Fellowship (SBG) and a Research Leave Fellowship (MEM). RCR received funding from the Wellcome Trust and BBSRC (UK) for related work.

Note: There were no conflicts of interest.


1. Gordon MA, Walsh AL, Chaponda M, Soko D, Mbvwinji M, Molyneux ME, et al. Bacteraemia and mortality among adult medical admissions in Malawi: predominance of nontyphi salmonellae and Streptococcus pneumoniae. J Infect 2001; 42:44–49.
2. Vugia DJ, Kiehlbauch JA, Yeboue K, N'Gbichi JM, Lacina D, Maran M, et al. Pathogens and predictors of fatal septicemia associated with human immunodeficiency virus infection in Ivory Coast, west Africa. J Infect Dis 1993; 168:564–570.
3. Archibald LK, den Dulk MO, Pallangyo KJ, Reller LB. Fatal Mycobacterium tuberculosis bloodstream infections in febrile hospitalized adults in Dar es Salaam, Tanzania. Clin Infect Dis 1998; 26:290–296.
4. Gilks CF, Brindle RJ, Otieno LS, Simani PM, Newnham RS, Bhatt SM, et al. Life-threatening bacteraemia in HIV-1 seropositive adults admitted to hospital in Nairobi, Kenya. Lancet 1990; 336:545–549.
5. Thamlikitkul V, Dhiraputra C, Paisarnsinsup T, Chareandee C. Nontyphoidal salmonella bacteraemia: clinical features and risk factors. Trop Med Int Health 1996; 1:443–448.
6. Gordon MA, Banda HT, Gondwe M, Gordon SB, Boeree MJ, Walsh AL, et al. Nontyphoidal salmonella bacteraemia among HIV-infected Malawian adults: high mortality and frequent recrudescence. AIDS 2002; 16:1633–1641.
7. Kedzierska K, Churchill M, Maslin CL, Azzam R, Ellery P, Chan HT, et al. Phagocytic efficiency of monocytes and macrophages obtained from Sydney blood bank cohort members infected with an attenuated strain of HIV-1. J Acquir Immune Defic Syndr 2003; 34:445–453.
8. Kedzierska K, Ellery P, Mak J, Lewin SR, Crowe SM, Jaworowski A. HIV-1 down-modulates gamma signaling chain of Fc gamma R in human macrophages: a possible mechanism for inhibition of phagocytosis. J Immunol 2002; 168:2895–2903.
9. Koziel H, Eichbaum Q, Kruskal BA, Pinkston P, Rogers RA, Armstrong MY, et al. Reduced binding and phagocytosis of Pneumocystis carinii by alveolar macrophages from persons infected with HIV-1 correlates with mannose receptor downregulation. J Clin Invest 1998; 102:1332–1344.
10. Koziel H, Li X, Armstrong MY, Richards FF, Rose RM. Alveolar macrophages from human immunodeficiency virus-infected persons demonstrate impaired oxidative burst response to Pneumocystis carinii in vitro. Am J Respir Cell Mol Biol 2000; 23:452–459.
11. Day RB, Wang Y, Knox KS, Pasula R, Martin WJ, Twigg HL III. Alveolar macrophages from HIV-infected subjects are resistant to Mycobacterium tuberculosis in vitro. Am J Respir Cell Mol Biol 2004; 30:403–410.
12. Reardon CC, Kim SJ, Wagner RP, Koziel H, Kornfeld H. Phagocytosis and growth inhibition of Cryptococcus neoformans by human alveolar macrophages: effects of HIV-1 infection. AIDS 1996; 10:613–618.
13. Gordon S. Alternative activation of macrophages. Nat Rev Immunol 2003; 3:13–35.
14. Monack DM, Bouley DM, Falkow S. Salmonella typhimurium persists within macrophages in the mesenteric lymph nodes of chronically infected Nramp1+/+ mice and can be reactivated by IFN{gamma} neutralization. J Exp Med 2004; 199:231–241.
15. Gulig PA, Doyle TJ, Clare-Salzler MJ, Maiese RL, Matsui H. Systemic infection of mice by wild-type but not Sp− Salmonella typhimurium is enhanced by neutralization of gamma interferon and tumor necrosis factor alpha. Infect Immun 1997; 65:5191–5197.
16. Vazquez-Torres A, Jones-Carson J, Mastroeni P, Ischiropoulos H, Fang FC. Antimicrobial actions of the NADPH phagocyte oxidase and inducible nitric oxide synthase in experimental salmonellosis. I. Effects on microbial killing by activated peritoneal macrophages in vitro. J Exp Med 2000; 192:227–236.
17. Kagaya K, Watanabe K, Fukazawa Y. Capacity of recombinant gamma interferon to activate macrophages for salmonella-killing activity. Infect Immun 1989; 57:609–615.
18. Gordon MA, Jack DL, Dockrell DH, Lee ME, Read RC. Gamma interferon enhances internalization and early nonoxidative killing of Salmonella enterica serovar Typhimurium by human macrophages and modifies cytokine responses. Infect Immun 2005; 73:3445–3452.
19. Janssen R, Van Wengen A, Verhard E, de Boer T, Zomerdijk T, Ottenhoff TH, et al. Divergent role for TNF-alpha in IFNgamma-induced killing of Toxoplasma gondii and Salmonella typhimurium contributes to selective susceptibility of patients with partial IFNgamma receptor 1 deficiency. J Immunol 2002; 169:3900–3907.
20. Kincy-Cain T, Clements JD, Bost KL. Endogenous and exogenous interleukin-12 augment the protective immune response in mice orally challenged with Salmonella dublin. Infect Immun 1996; 64:1437–1440.
21. Mastroeni P, Harrison JA, Robinson JH, Clare S, Khan S, Maskell DJ, et al. Interleukin-12 is required for control of the growth of attenuated aromatic-compound-dependent salmonellae in BALB/c mice: role of gamma interferon and macrophage activation. Infect Immun 1998; 66:4767–4776.
22. Altare F, Lammas D, Revy P, Jouanguy E, Doffinger R, Lamhamedi S, et al. Inherited interleukin 12 deficiency in a child with bacille Calmette–Guerin and Salmonella enteritidis disseminated infection. J Clin Invest 1998; 102:2035–2040.
23. de Jong R, Altare F, Haagen IA, Elferink DG, Boer T, van Breda Vriesman PJ, et al. Severe mycobacterial and salmonella infections in interleukin-12 receptor-deficient patients. Science 1998; 280:1435–1438.
24. Arai T, Hiromatsu K, Nishimura H, Kimura Y, Kobayashi N, Ishida H, et al. Effects of in vivo administration of anti-IL-10 monoclonal antibody on the host defence mechanism against murine salmonella infection. Immunology 1995; 85:381–388.
25. Casado JL, Navas E, Frutos B, Moreno A, Martin P, Hermida JM, et al. Salmonella lung involvement in patients with HIV infection. Chest 1997; 112:1197–1201.
26. Fernandez GM, Ramos JM, Nunez A, de Gorgolas M. Focal infections due to nontyphi salmonella in patients with AIDS: report of 10 cases and review. Clin Infect Dis 1997; 25:690–697.
27. Mankhambo LA, Chiwaya KW, Phiri A, Graham SM. Lobar pneumonia caused by nontyphoidal salmonella in a Malawian child. Pediatr Infect Dis J 2006; 25:1190–1192.
28. Gordon SB, Irving GR, Lawson RA, Lee ME, Read RC. Intracellular trafficking and killing of Streptococcus pneumoniae by human alveolar macrophages are influenced by opsonins. Infect Immun 2000; 68:2286–2293.
29. Porcheray F, Samah B, Leone C, Dereuddre-Bosquet N, Gras G. Macrophage activation and human immunodeficiency virus infection: HIV replication directs macrophages towards a pro-inflammatory phenotype while previous activation modulates macrophage susceptibility to infection and viral production. Virology 2006; 349:112–120.
30. Giri MS, Nebozhyn M, Showe L, Montaner LJ. Microarray data on gene modulation by HIV-1 in immune cells: 2000–2006. J Leukoc Biol 2006; 80:1031–1043.
31. Kandil O, Fishman JA, Koziel H, Pinkston P, Rose RM, Remold HG. Human immunodeficiency virus type 1 infection of human macrophages modulates the cytokine response to Pneumocystis carinii. Infect Immun 1994; 62:644–650.
32. Bergamini A, Faggioli E, Bolacchi F, Gessani S, Cappannoli L, Uccella I, et al. Enhanced production of tumor necrosis factor-alpha and interleukin-6 due to prolonged response to lipopolysaccharide in human macrophages infected in vitro with human immunodeficiency virus type 1. J Infect Dis 1999; 179:832–842.
33. Bergamini A, Bolacchi F, Bongiovanni B, Colizzi V, Cappelli G, Uccella I, et al. Human immunodeficiency virus type 1 infection modulates the interleukin (IL)-1beta and IL-6 responses of human macrophages to CD40 ligand stimulation. J Infect Dis 2000; 182:776–784.
34. Muller F, Aukrust P, Lien E, Haug CJ, Froland SS. Enhanced interleukin-10 production in response to Mycobacterium avium products in mononuclear cells from patients with human immunodeficiency virus infection. J Infect Dis 1998; 177:586–594.
35. Stylianou E, Aukrust P, Kvale D, Muller F, Froland SS. IL-10 in HIV infection: increasing serum IL-10 levels with disease progression: down-regulatory effect of potent antiretroviral therapy. Clin Exp Immunol 1999; 116:115–120.
36. Chehimi J, Starr SE, Frank I, D'Andrea A, Ma X, MacGregor RR, et al. Impaired interleukin 12 production in human immunodeficiency virus-infected patients. J Exp Med 1994; 179:1361–1366.
37. Marshall JD, Chehimi J, Gri G, Kostman JR, Montaner LJ, Trinchieri G. The interleukin-12-mediated pathway of immune events is dysfunctional in human immunodeficiency virus-infected individuals. Blood 1999; 94:1003–1011.
38. Twigg HL III, Spain BA, Soliman DM, Knox K, Sidner RA, Schnizlein-Bick C, et al. Production of interferon-gamma by lung lymphocytes in HIV-infected individuals. Am J Physiol 1999; 276(2 Pt 1):L256–L262.
39. Wood KL, Knox KS, Wang Y, Day RB, Schnizlein-Bick C, Twigg HL III. Apoptosis of CD57+ and CD57− lymphocytes in the lungs and blood of HIV-infected subjects. Clin Immunol 2005; 117:294–301.
40. Lee C, Tomkowicz B, Freedman BD, Collman RG. HIV-1 gp120-induced TNF-{alpha} production by primary human macrophages is mediated by phosphatidylinositol-3 (PI-3) kinase and mitogen-activated protein (MAP) kinase pathways. J Leukoc Biol 2005; 78:1016–1023.
41. Li JC, Lee DC, Cheung BK, Lau AS. Mechanisms for HIV Tat upregulation of IL-10 and other cytokine expression: kinase signaling and PKR-mediated immune response. FEBS Lett 2005; 579:3055–3062.
42. Isler P, de Rochemonteix BG, Songeon F, Boehringer N, Nicod LP. Interleukin-12 production by human alveolar macrophages is controlled by the autocrine production of interleukin-10. Am J Respir Cell Mol Biol 1999; 20:270–278.
43. Pie S, Truffa-Bachi P, Pla M, Nauciel C. Th1 response in Salmonella typhimurium-infected mice with a high or low rate of bacterial clearance. Infect Immun 1997; 65:4509–4514.

AIDS; bacteria; cytokines; human; monocytes/macrophages

© 2007 Lippincott Williams & Wilkins, Inc.