Recent insights in the development and functions of insect hemocytes

Abstract

Hemocytes mediate defense responses that are collectively referred to as cellular immunity. Here I highlight recent progress in the study of hemocyte ontogeny and function. Studies conducted primarily in D. melanogaster, mosquitoes and Lepidoptera identify a number of transcription factors and signaling pathways that regulate hemocyte differentiation and proliferation. Recent single cell RNA sequencing studies identify genes that distinguish transitional states in different hemocyte populations and changes that occur during immune responses. Several new studies expand understanding of the functions of sessile hemocytes. Recent results also enhance understanding of the signaling pathways that regulate hemocyte functions.

Introduction

A wide range of microbes and multicellular parasites infect insects, which defend themselves through an innate immune system that consists of physical barriers, cells, and different types of molecules. Cuticle serves as a protective outer barrier and also lines the tracheal system, foregut, and hindgut. The midgut, which is the primary site of digestion, produces reactive oxygen species and other defense molecules while internal organs like the fat body produce defense molecules like antimicrobial peptides (AMPs) that are secreted into the hemolymph that fills the body cavity (hemocoel). Hemolymph consists of a fluid component, plasma, and hemocytes which are the immune cells of insects. Hemocyte-mediated defense responses are collectively referred to as cellular immunity. Cellular immune responses include the phagocytosis, nodulation, or encapsulation of pathogens by hemocytes (Figure 1). Hemocytes additionally secrete defense molecules that kill pathogens, express a range of receptors and cytokines, and participate in coagulation responses that form clots (Figure 1). Several reviews have been published on hemocytes and cellular immunity. These include a comparative summary we wrote in 2002 that was updated in 2008, and a review of the mosquito cellular immunity literature in 2014 [1–3]. Many additional findings have since been reported. Most of these new results derive from studies of Drosophila melanogaster, mosquitoes and Lepidoptera but also include insects in select other taxa. Here, I highlight areas of progress by emphasizing primary articles published since 2015. I also cite some more recent reviews to accommodate space constraints.

Figure 1:

Cellular defense responses, humoral effector molecules and regulators of hemocyte functions in insects. In the center of the figure is a stylized image of an insect hemocyte. The arrow to the left points to the four major cellular defenses that are mediated by hemocytes. The arrow to the right points to five of the most important humoral effector molecules hemocytes produce. The upper arrow points to key receptors with functions in regulating hemocyte-mediated cellular defenses, while the lower arrow points to cytokines that regulate hemocyte functions or communication between hemocytes and other tissues. Small case statements below some headings provide brief explanations of terms while additional details are presented in the main text.

Hemocyte types

Studies dating to the 1970s subdivide insect hemocytes into populations that are distinguished by morphology, function and a small number of molecular markers [1–3]. Naming of these populations varies somewhat between taxa (Figure 2). In D. melanogaster larvae, for example, progenitor cells in hematopoietic organs are named prohemocytes while hemocyte types in circulation with functions in cellular immunity are phagocytic plasmatocytes, capsule-forming lamellocytes, and melanin-producing crystal cells [5–7]. One subgroup of drosophilids have capsule-forming hemocytes named nematocytes that morphologically differ from lamellocytes while another subgroup of drosophilids have capsule-forming hemocytes named multinucleated giant hemocytes that are distinct from lamellocytes and nematocytes [8,9]. In larval stage Lepidoptera, progenitor, phagocytic, and melanin-producing hemocytes are named prohemocytes, granulocytes and oenocytoids respectively, capsule-forming hemocytes are named plasmatocytes and a fifth distinct cell type that potentially transports cuticle components are named spherule cells [1]. In adult mosquitoes, putatively progenitor, phagocytic, and melanin-producing hemocytes are named like Lepidoptera, whereas no capsule-forming hemocytes or spherule cells are present [3,4,10]. Recent studies of hemocytes in bees, beetles, crickets and true bugs also use naming schemes that are similar to Lepidoptera [11–14]. Other types of hemocytes have also been described in different types of insects [1,3,7]. As discussed below, some of these are distinct from the main hemocyte types listed above. In contrast, others are likely contaminants that have been erroneously referred to as hemoctyes. Examples of this are thrombocytoids and adipohemocytes which have been described as hemocyte types in adult mosquitoes but are likely pericardial cells from the dorsal vessel (heart) and fat body respectively that are easily dislodged when collecting hemolymph from adults [15].

Figure 2:

Names of the hemocytes present in Drosophila melanogaster (family Drosophilidae, Order Diptera), moths and butterflies (order Lepidoptera), and mosquitoes (family Culicidae, order Diptera). Primary functions of the hemocyte types are indicated to the left. Most studies of hemocytes in D. melanogaster and Lepidoptera focus on the larval stage while most studies in mosquitoes focus on the adult stage. Similar coloration for the phagocytic hemocytes (plasamatocytes or granulocytes), and melanin-producing hemocytes (crystal cells and oenocytoids) stem from shared expression of genes that suggest common origins. Lamellocytes in D. melanogaster and lepidopteran plasmatocytes currently exhibit dissemilarities, while insufficient information exists to determine if hemocytes named prohemocytes in different species share a common orgin (see main text). Recently identified hemocyte types like primoctyes in D. melanogaster and megacytes in mosquitoes are not shown (see main text). Capsule-forming nematocytes and multinucleated giant hemocytes identified in drosophilids outside of D. melanogaster are also not shown (see main text).

Hemocyte ontogeny

Studies beginning in the early 2000s showed that hemocytes in larval stage D. melanogaster derive from two sources: 1) progenitor cells called prohemocytes that derive from embryonic mesoderm and reside in paired hematopoietic organs along the dorsal vessel (insect heart), and 2) a lineage of self-renewing plasmatocytes in the hemocoel that also derive from embryonic mesoderm [6,16,17]. Hemocytes in larval stage Lepidoptera similarly derive from prohemocytes that reside in paired hematopoietic organs in the thorax and granulocytes (=granular cells) of embryonic origin in the hemocoel that self-renew [18–20]. In mosquitoes, hemocytes have primarily been studied in adults where no hematopoietic organs have been identified. Granulocytes comprise 90–95% of the cells in the hemocoel, which rapidly proliferate in response to blood feeding or infection by microbes like bacteria, while oenocytoids and prohemocytes are much less abundant [21,22]. Hemocytes also proliferate in response to immune challenge in adult orthopterans like Locusta migratoria [23]. More recently, studies in D. melanogaster show that prohemocytes in hematopoietic organs differentiate into all of the hemocyte types in hemolymph, while self-renewing plasmatocytes from the embryonic lineage can differentiate into crystal cells and lamellocytes [6,24,25]. A number of transcription factors (Serpent, Lozenge, Glial cell missing) and signaling pathways (Notch, Toll, Jak-STAT, Jun kinase, Ras-MAPK) have been identified that regulate these differentiation events [6,26–29). Insulin-TOR and Ras-MAPK pathways have also been shown to regulate the proliferation of hemocytes in D. melanogaster, mosquitoes and Lepidoptera [6,21,30–33].

Recent single-cell RNA sequencing (scRNA-seq) studies together with other approaches affirm that the previously mentioned major hemocyte types previously identified in D. melanogaster larvae and adult mosquitoes form unique hemocyte clusters that express distinct suites of genes [7]. These studies also advance the field by identifying gene markers that distinguish transitional states between hemocyte types during differentiation while providing insights about homologous relationships between hemocyte types in different insect species [7]. In the case of D. melanogaster larvae, scRNA-seq and single-cell mass cytometry identify changes in state of plasmatocytes, crystal cells, and lamellocytes under different inflammatory conditions as well as previously unknown subsets of lamellocytes and crystal cells that express fibroblast growth factor-like ligands that are required for defense against parasitoid wasps [34–36]. While mitotically active hemocytes in circulation have been suggested to represent a prohemocyte population in D. melanogaster larvae [36], other evidence points to prohemocyte clusters being restricted to hematopoietic organs [7]. A rare cluster of cells in hematopoietic organs and in circulation have also been identified by scRNA-seq that represents a previously unknown hemocyte class with unclear functions named primocytes [7,36]. Lastly, scRNA-seq data indicate all hemocyte types in D. melanogaster larvae persist after pupation [37]. ScRNA-seq in combination with other methods similarly reveal genes that distinguish prohemocytes, granulocytes and oenocytoids in adult Anopheles gambiae and Aedes aegypti mosquitoes and associated transitional states [38–41]. ScRNA-seq data followed by functional experiments further show that activation of the Toll pathway after infection by malaria parasites (Plasmodium) stimulates some granulocytes in An. gambiae to differentiate into a previously unknown hemocyte type named megacytes which have anti-Plasmodium functions [42]. ScRNA-seq studies have not been used to characterize hemocytes in other insects although changes in hemocyte state were identified by scRNA-seq in Bombyx mori larvae (Lepidoptera) after baculovirus infection [43].

Comparing between species indicates D. melanogaster crystal cells, oenocytoids in B. mori [43], and oenocytoids in some mosquito studies [38,39] express similar clusters of marker genes [7]. Comparing gene markers in phagocytic plasmatocytes from D. melanogaster to gene markers for phagocytic granulocytes in B. mori [43] and phagocytic granulocytes in some mosquito studies [38,39] also share common features [7]. In contrast, only a few markers that identify capsule-forming lamellocytes from D. melanogaster are shared with capsule-forming plasmatocytes in B. mori, which currently makes it unclear if these hemocytes share a common origin [7]. Lamellocyte markers are also not shared with any mosquito hemocyte types but this finding is consistent with mosquito hemocytes also being unknown to form capsules [4,7,10]. Cells identified as prohemocytes in different insect species show a few similarities in marker genes but additional study is still needed to determine if they share common origins or not.

Circulating and sessile hemocytes

Circulating hemocytes have historically been the focus of cellular immunity studies [1–5] but in recent years increased attention has focused on factors that cause hemocytes in circulation to attach to surfaces in the hemocoel and become sessile or induce sessile hemocytes to reenter circulation [44,45]. Several studies have also provided new information on the distribution of sessile hemocytes in the hemocoel of insects and their functions. In the absence of immune challenge, sessile hemocytes in larval stage D. melanogaster are most abundant in segmentally iterated domains named sessile hemocyte compartments (SHCs) [16]. Some results indicate SHCs function as hematopoietic compartments in addition to the hematopoietic organs [25]. The close proximity of neurons of the peripheral nervous system to SHCs have also been suggested to play a role in recruiting circulating hemocytes to SHCs [46]. The Nimrod transmembrane receptor Eater on plasmatocytes mediates phagocytosis of Gram-positive bacteria (Figure 1) but is also involved in binding to SHCs via interactions with the Multiplexin which is a Collagen XV/VVIII ortholog [47,48]. Recent results additionally identify transmembrane proteins named ninjurins as adhesion molecules that direct hemocytes to SHCs, while ecdysteroids and transcription factors that downregulate ninjurins result in sessile hemocytes reentering circulation before pupation [49]. Ninjurins are also markers of phagocytic granulocytes in mosquitoes [40].

Sessile hemocytes occur throughout the hemocoel in larval stage mosquitoes, but bacterial infection induces hemocytes to preferentially aggregate in tracheal tufts near the incurrent opening to the dorsal vessel (=heart) [50]. Sessile hemocytes in adult stage mosquitoes and insects in several other orders preferentially bind to ostia which are paired valves where hemolymph enters the dorsal vessel that pumps hemolymph [51]. Hemocyte aggregation around ostia increases in response to infection by bacteria, but this increase is reduced by RNA interference (RNAi) mediated knockdown of genes encoding extracellular matrix proteins, complement-like proteins and a transglutaminase [52–54]. Nitric oxide produced by hemocytes (Figure 1) adhering to ostia also modulate heart rate, which potentially promotes clearance of bacteria by altering hemolymph flow rates through the dorsal vessel [55]. Reciprocally, studies in Lepidoptera document sessile hemocytes rapidly entering circulation following infection by pathogens or injection of pathogen derived cell surface molecules like fungal ß glucans [56,57], which are recognized by ß glucan recognition proteins (ßGRPs) (Figure 1).

Other advances focus on D. melanogaster embryos where plasmatocytes are persistently sessile and migrate to wound sites. The earliest signal released from wound sites that attracts plasmatocytes is hydrogen peroxide generated by an NADPH oxidase (DUOX) which is activated by calcium that is putatively released from wounded cells [58]. A complement like protein named Macroglobulin complement-related was also recently identified as a wound-induced chemoattractant that operates in conjunction with hydrogen peroxide [59]. Circulating hemocytes in D. melanogaster larvae can also contribute to coagulation and wound closure (Figure 1) by binding to wound sites [44]. Studies in larval stage Lepidoptera identified a peptide that stimulates chemotactic movement by phagocytic granulocytes and wound sealing that is produced by epidermal cells [60]. Lepidoptera also produce cytokines named plasmatocyte spreading peptide (PSP) (=growth blocking peptide (GBP)) (Figure 1) which is produced by granulocytes and the fat body, and rapidly stimulates plasmatocytes to adhere to foreign surfaces as occurs during encapsulation [61,62]. A PSP/GFP-like peptide in D. melanogaster activates cellular and humoral defense responses through effects on multiple signaling pathways [63,64]. Recent studies demonstrate sessile plasmatocytes in D. melanogaster embryos infiltrate the germ band and phagocytize apoptotic cells [65]. In subsequent life stages, plasmatocytes infiltrate the fat body and the ovaries where they clear cellular debris from follicle cells [66–68]. Infection by bacteria can also induce plasmatocyte entry into the brain [69].

Signaling and effector functions

Insects hemocytes detect pathogens, parasitoids, and damaged cells or tissues through cell surface receptors that also regulate hemocyte-mediated defense responses like phagocytosis and encapsulation [70, 71]. Opsonins, which are secreted proteins that bind to different types of foreign targets, also promote recognition by hemocytes [71]. Key cell surface receptors implicated in recognition of different foreign targets and apoptotic cells include several Nimrod family members like Eater, certain integrin family members, and the CD36-like factor Santa Maria, while C-type lectins, thioester-containing (complement-like) proteins, peptidoglycan recognition proteins (PGRPs), and ßGRBPs that are glucan sensors (Figure 1). The above factors additionally function as pattern recognition receptors (PRRs) which together with cytokines activate immune signaling pathways such as Toll, immune-deficiency (Imd), Jun-kinase (JNK), and Janus-kinase-signal transducers and activators of transcription (JAK-STAT) [70, 71]. The Toll and Imd pathways are best known for regulating AMP expression upon recognition of bacteria and fungi but also have other functions including antiviral defense [72]. In the case of Toll signaling, recognition of peptidoglycan by PGRPs activates serine protease cascades which processes precursor proteins (pro-Spatzle family members) into cytokines (Spatzle family members) that bind to and activate Toll family receptors [73]. This is in contrast to Toll-like receptors (TLRs) in vertebrates which function as PRRs by directly binding microbial products like lipopolysaccharide (LPS). However, recent results in Lepidoptera show that one Toll family member (Toll-9) functions as a PRR that also recognizes LPS through partnering with a coreceptor while Toll-5 recognizes baculoviruses by binding a virion envelope component [74, 75]. JNK pathway activation can occur through a receptor named Grindewald in D. melanogaster that binds a Tumor Necrosis Factor-like cytokine (Eiger) or Platelet-Derived Growth Factor (PDGF)/Vascular Endothelial Growth Factor (VEGF)-related receptor (Pvr) that binds cytokines named Pvf ligands [76]. The JNK pathway upregulates the expression of a number of immune genes in insects while also having functions in cellular defenses like phagocytosis in aphids [71,76,77]. Pvf ligands also serve as chemotactic signals that regulate plasmatocyte migration in D. melanogaster embryos [78]. Leptin-like proteins in the unpaired family (Upd1–3) function as cytokines that activate the JAK-STAT pathway by binding to a transmembrane receptor named Domeless [71]. In turn, the JAK-STAT pathway regulates responses that include defense against viruses and parasitoid wasps, wound healing, activation of Turandot genes that confer stress protection against antimicrobial peptides, and behavioral responses that occur during infection [5,70,71,79,80].

The hemocyte types that mediate cellular immune responses like phagocytosis, encapsulation, melanization, and coagulation have been known for decades (Figure 1, 2). Most of the receptors, cytokines, and signaling pathways that regulate cellular immune responses were also identified before 2014 (Figure 1). However, recent studies have further advanced the field by revealing previously uncharacterized linkages between pathways. For example, the serine protease cascades that recognize microbes and activate the Toll pathway by processing pro-Spaetzle are now known to also regulate processing of prophenoloxidases (PPOs) in melanin-producing oenocytoids (Lepidoptera) or crystal cells (D. melanogaster) [81,82]. Results also show that crystal cells interact with the tracheal system and use PPO2 to transport oxygen, which identifies a previously unknown role for melanin-producing hemocytes and the circulatory system in respiration [83]. Recent advances further identify important roles for hemocytes in inter-tissue communication. In D. melanogaster, hemocytes release Upd ligands that activate the JAK/STAT pathway and Pvf ligands that activate JNK signaling in other tissues [84,85]. Ablation of hemocytes in larval stage D. melanogaster is also fatal for development of pupae into adults due to functions that control the microbiota, clear dead cells, and are required for developmental reprogramming [86].

Lastly, recent studies indicate insects exhibit immune priming responses where prior exposure to sublethal doses of a pathogen enhances resistance to subsequent infection [87]. Immune priming is mediated by nonspecific immune mechanisms, which include persistent increases in the number of hemocytes in circulation and titers of several antimicrobial molecules following pathogen exposure [88,89]. Priming responses can also be trans-generationally transmitted [90–92]. One mechanism identified in D. melanogaster that primes increased defense is olfactory detection of parasitoid wasps which activates processes that stimulate lamellocyte formation [93]. Bacterial symbionts in Tsetse flies and the microbiota in larval stage Drosophila also up-regulate the expression of odorant binding proteins (OBPs) with functions in hematopoiesis [94], while other factors like the adipokine NimrodB5 have been identified that downregulate hemocyte proliferation in response to changes in metabolic state like food limitation [95]. These findings indicate insects can perceive stimuli that are associated with increased risks of infection and also regulate metabolic processes that affect priming responses and host fitness [96]. In contrast, while some studies suggest insects may also develop immune memory [97,98], experimental support for underlying molecular mechanisms are currently lacking.

Concluding remarks

In this short review, I call attention to recent studies that have improved understanding of hematopoiesis, the origins and functions of sessile hemocytes, and the signaling pathways that regulate cellular immune responses. I also call attention to the roles of hemocytes in inter-tissue communication, functions that extend beyond immunity, and priming responses. Future research is certain to further build on recent findings. Better understanding of inter-tissue communication during immune responses, the processes that regulate priming responses, and the non-immune functions of hemocytes during metamorphosis are all likely to see further advances in the near future. The long-standing issue of whether insects or related invertebrates like Crustacea develop immune memory and, if so, underlying molecular mechanisms also requires more study.