Proteophosophoglycans regurgitated by Leishmania-infected sand flies target the L-arginine metabolism of host macrophages to promote parasite survival

Abstract

All natural Leishmania infections start in the skin; however, little is known of the contribution made by the sand fly vector to the earliest events in mammalian infection, especially in inflamed skin that can rapidly kill invading parasites. During transmission sand flies regurgitate a proteophosphoglycan gel synthesized by the parasites inside the fly midgut, termed promastigote secretory gel (PSG). Regurgitated PSG can exacerbate cutaneous leishmaniasis. Here, we show that the amount of Leishmania mexicana PSG regurgitated by Lutzomyia longipalpis sand flies is proportional to the size of its original midgut infection and the number of parasites transmitted. Furthermore, PSG could exacerbate cutaneous L. mexicana infection for a wide range of doses (10-10,000 parasites) and enhance infection by as early as 48 hours in inflamed dermal air pouches. This early exacerbation was attributed to two fundamental properties of PSG: Firstly, PSG powerfully recruited macrophages to the dermal site of infection within 24 hours. Secondly, PSG enhanced alternative activation and arginase activity of host macrophages, thereby increasing L-arginine catabolism and the synthesis of polyamines essential for intracellular parasite growth. The increase in arginase activity promoted the intracellular growth of L. mexicana within classically activated macrophages, and inhibition of macrophage arginase completely ablated the early exacerbatory properties of PSG in vitro and in vivo. Thus, PSG is an essential component of the infectious sand fly bite for the early establishment of Leishmania in skin, which should be considered when designing and screening therapies against leishmaniasis.

Figure 1. Transmitted L. mexicana dose in relation to amount of PSG regurgitated by sand flies.

(A) Relationship between numbers of L. mexicana egested by Lu. longipalpis sand flies collected by capillary feeding and size of the sand fly midgut parasite load quantified by direct counting immediately after capillary feeding. (B) Dot-blots of L. mexicana PSG standards and capillary feeds from L. mexicana-infected sand flies probed with anti-PSG sera or normal rabbit serum (NRS). (C) Area under curve analysis of digitized PSG standards (inset) reveals high and low egestion of PSG. (D&E) Relationship between the total midgut parasite load and the amount of PSG egested. (F&G) Relationship between the transmitted dose of parasites and the amount of PSG egested. **, P<0.005.

Figure 2. PSG exacerbates low and high dose L. mexicana infections in mice.

(A–C) Co- inoculation of L. mexicana PSG (1 µg) with 10 (A), 100 (B) and 10,000 (C) L. mexicana metacyclic promastigotes in the rump of BALB/c mice (6 mice per group) resulted in earlier onset of lesion appearance and faster lesion evolution, displaying higher final parasite burdens (shown in Table 1). *, P<0.05; **, P<0.005.

Figure 3. PSG recruits macrophages to the site of transmission and enhances their infection with L. mexicana.

(A–C) Cellular content within air pouches inflated on the backs of BALB/c mice. Seventy two hour kinetic of neutrophil (A) and macrophage (B) recruitment in response to 1×10³ L. mexicana metacyclic promastigotes co-injected with 1 µg L. mexicana PSG and 1 µg Lu. longipalpis saliva. (C) Neutrophil and macrophage content of air pouches in response to 1 µg PSG, 1 µg sand fly saliva or 1 µg of PSG and saliva after 24 hours. Cell type was determined by morphology of at least 200 Giemsa-stained cells per air pouch (4–6 mice per group). Data representative of duplicate (A&B) or triplicate (C) experiments. (D) Macrophages recovered from 48 hour PSG -loaded inflammatory air pouches were infected ex vivo at a MOI of 1∶1 for a further 48 hours. Inflammatory air pouches were generated by the injection of 200 U INFγ and 500 U TNFα 18 hours prior to the injection of PSG. Parasite burden (D) was determined by microscopy of at least 200 Giemsa-stained macrophages in quadruplicate. Representative experiment of duplicates is shown. Unless they are linked with a bar all test groups are compared to their relevant saline control; *, P<0.05; **, P<0.005; ***, P<0.0005.

Figure 4. Co-incubation of L. mexicana with PSG enhances BALB/c macrophage infections in vitro under a range of activation states.

(A–D) Kinetic of parasite growth in infected BMMΦ (5×10⁴; MOI 5∶1) in the presence or absence of classical activators and vector-derived products. (A) Control L. mexicana infections of PBS (unstimulated, open circles), IFNγ and TNFα-treated (classically activated: CAMΦ, open squares) or IL-4-treated (alternatively activated: AAMΦ, open diamonds) macrophages. The role of L. mexicana PSG (0.25 µg) (B), Lu. longipalpis sand fly saliva (0.25 µg) (C) or a combination of PSG and saliva (0.25 µg of each) (D) (closed diamonds) was assessed for infection in CAMΦ. (E&F) 48 hour parasite burden of unstimulated BMMΦ, AAMΦ and CAMΦ infected in the presence or absence of 0.25 µg PSG macrophages was assessed by microscopy of infected cells (E), or by transformation assay of amastigotes released from macrophages (F). Growth was determined by direct counting of transformed promastigotes by haemocytometer in triplicate. (G–I) Effect of PSG (0.25 µg), saliva (0.25 µg), PSG and saliva (0.25 µg of each), deglycosylated (degly) PSG (0.25 µg) or saline on LPG2^−/− (phosphoglycan-deficient) L. mexicana infections of unstimulated (G), AA (H) and CA (I) macrophages. Amastigote burden of infected macrophages determined by microscopy of at least 200 Giemsa-stained cells in triplicate or quadruplicate. Data representative of triplicate experiments. Unless they are linked with a bar all test groups are compared to their relevant saline control; ns, not significant P>0.05;*, P<0.05; **, P<0.005; ***, P<0.0005.

Figure 5. Exposure to PSG increases the alternative activation and arginase activity of macrophages.

(A) 5×10⁵ unstimulated, alternatively activated and classically activated BALB/c BMMΦ were incubated in the presence or absence of 1 µg L. mexicana PSG for 48 hours. Western blot of macrophage lysates (3 µg protein/lane) were probed for murine Arginase-1, Ym-1, iNOS and β-actin was used as loading control. (B&C) Arginase assay of L. mexicana-infected (B) or uninfected (C) macrophage lysates (unstimulated, AA or CA). (D) NO assay of infected macrophage culture supernatants. Macrophages were co-incubated with 1 µg L. mexicana PSG, 1 µg Lu. longipalpis sand fly saliva or 1 µg PSG and saliva for 48 hours. Data is representative of duplicate (A) or triplicate (B–D) experiments and normalised to the average unistimulated+saline control macrophage arginase/iNOs activity. Unless they are linked with a bar all test groups are compared to their relevant saline control; ns, not significant P>0.05; *, P<0.05; **, P<0.005; ***, P<0.0005.

Figure 6. PSG promotes L. mexicana infection of mice by increasing the arginase activity of macrophages at the site of infection.

(A) Parasite burden of BALB/c air pouches infected with 1×10⁶ L. mexicana metacyclic promastigotes ±1 µg L. mexicana PSG, ±100 µg nor-NOHA, ±100 µg L-ornithine for 48 hours. Macrophages were recovered from infected air pouches and separated by adherence to plastic. (B&C) Effect of vector-derived products on macrophage parasite burden following arginase inhibition with nor-NOHA. BALB/c BMMΦ were infected with L. mexicana metacyclic promastigotes at a MOI 1∶1 in the presence of ±1 µg PSG, ±1 µg Lu. longipalpis saliva or PSG and saliva; ±100 µM nor-NOHA; ±100 µM (B); or L-ornithine (C) for 48 hours. Viable amastigote burdens of macrophages infected in vivo (A), or in vitro (B&C) was determined by transformation assay of amastigotes liberated from either 2.5×10⁵ air pouch macrophages (4 or 8 mice per group) (A), or from 5×10⁵ in vitro infected BMMΦ (B&C). Data is pooled from duplicate experiments (A), or representative of duplicate experiments (B&C). Unless they are linked with a bar all test groups are compared to their relevant saline control; ns, not significant P>0.05; *, P<0.05; **, P<0.005; ***, P<0.0005.