Evidence for a "wattle and daub" model of the cyst wall of entamoeba

Abstract

The cyst wall of Entamoeba invadens (Ei), a model for the human pathogen Entamoeba histolytica, is composed of fibrils of chitin and three chitin-binding lectins called Jacob, Jessie3, and chitinase. Here we show chitin, which was detected with wheat germ agglutinin, is made in secretory vesicles prior to its deposition on the surface of encysting Ei. Jacob lectins, which have tandemly arrayed chitin-binding domains (CBDs), and chitinase, which has an N-terminal CBD, were each made early during encystation. These results are consistent with their hypothesized roles in cross-linking chitin fibrils (Jacob lectins) and remodeling the cyst wall (chitinase). Jessie3 lectins likely form the mortar or daub of the cyst wall, because 1) Jessie lectins were made late during encystation; 2) the addition to Jessie lectins to the cyst wall correlated with a marked decrease in the permeability of cysts to nucleic acid stains (DAPI) and actin-binding heptapeptide (phalloidin); and 3) recombinant Jessie lectins, expressed as a maltose-binding proteins in the periplasm of Escherichia coli, caused transformed bacteria to agglutinate in suspension and form a hard pellet that did not dissociate after centrifugation. Jessie3 appeared as linear forms and rosettes by negative staining of secreted recombinant proteins. These findings provide evidence for a "wattle and daub" model of the Entamoeba cyst wall, where the wattle or sticks (chitin fibrils likely cross-linked by Jacob lectins) is constructed prior to the addition of the mortar or daub (Jessie3 lectins).

Figure 1. Three-dimensional high-resolution fluorescence microscopy shows that Jacob lectins and chitin are each synthesized in discrete vesicles and deposited onto the parasite surface early during Ei encystation.

Parasites were fixed and permeabilized with nonionic detergent prior to labeling with antibodies to Jacob lectins (green), wheat germ agglutinin (red) that binds chitin, and DAPI (blue) that stains nuclei. Because encystation is not well-synchronized, in each case multiple labels were used on the same set of organisms in order to determine the order of events during encystation. Jacob lectins and chitin, which are absent in trophozoites (not shown), are each present in separate vesicles after 12 hrs of encystation. At 24 hrs there is substantially more Jacob lectin than chitin on the protist surface, while both Jacob and chitin are present in the wall of Ei encysting for 36 hrs. Bar is 10 microns. (A) is a composite of multiple optical sections, while (B and C) are each a single optical section.

Figure 2. Three-dimensional high-resolution fluorescence microscopy shows Jacob lectins and chitinase are expressed early during encystation of Ei in vitro, while Jessie lectins are expressed late during encystation.

Ei were allowed to encyst for 12 hrs (A to D), 24 hrs (E to H), 36 hrs (I to L), 48 hrs (M to P), or 72 hrs (Q to T) before they were fixed, permeabilized with non-ionic detergent, and then directly labeled with red anti-Jacob antibodies, green anti-chitinase antibodies, and blue anti-Jessie3 antibodies. At each time point, the same organism is shown with the three labels, and the merged three-color images are shown in the right hand-column. Not shown are control Ei trophozoites, which do not bind antibodies to Jacob, chitinase, or Jessie3. Also not shown are negative controls with non-immune rabbit antibodies, which did not bind to encysting organisms. Jacob lectins form large vesicles early during encystation; Jacob is the first lectin to appear on the surface of encysting Ei; and Jacob is a major component of the mature cyst wall. Chitinase appears next and is present in large vesicles that rarely overlap with those of Jacob lectins. In contrast to Jacob, most of the chitinase does not remain on the surface of mature cysts. Vesicles containing Jessie3 lectins appear late, and for the most part Jessie3 lectins are targeted to the cyst wall. Bar is 10 microns. Each image is a composite of multiple optical sections.

Figure 3. Three-dimensional high-resolution fluorescence microscopy shows that Jessie lectins (red) arrive late but eventually cover the surface of cysts.

Unlike encysting Ei in Figs. 1, 2, and 4, protists here were labeled with antibodies to Jessie3 prior to fixation, so that only the surfaces of the encysting parasites are labeled. While there is very little Jessie3 lectin on the parasite surface after 36 hrs (A), Jessie3 appears in increasing number of punctate spots, sometimes in linear arrays (arrows), in the cyst wall at 48 hrs (B) and 60 hrs (C). In contrast, at 72 hrs (D), Jessie3 has a swirling appearance in the Ei cyst wall. After Jessie3 is fully incorporated into the wall (D), the cyst becomes impermeable to DAPI, so that nuclei are no longer visible or are weakly visible. Bar is 10 microns. Each image is the composite of multiple optical sections.

Figure 4. Three-dimensional high-resolution fluorescence microscopy shows that coincident with the addition of Jessie lectins (red), cyst walls become impermeable to DAPI (blue) and phalloidin (green).

(A to C) After 72 hrs encystation, some Ei are still ameboid in appearance, some have relatively little Jessie3 lectin in the wall, and some protists have abundant Jessie3. DAPI and phalloidin stain well the immature cysts (Imm.) but fail to penetrate mature cysts. (D to F) In contrast, DAPI and phalloidin penetrate all cysts that have been frozen and thawed prior to staining with these reagents. Bar is 10 microns. Each image is a single optical section.

Figure 5. Chitin-binding is associated with the N-terminal Cys-rich domain of Jessie3.

Left panel: Coomassie blue stained SDS-PAGE shows an MBP-fusion-protein containing the CBD of Jessie3, which was purified with an amylose resin. A lower mol wt band is likely MBP alone, as only the MBP-Jessie3 CBD fusion-protein binds to chitin (right lane). Left center panel: A Coomassie-stained MBP-fusion-protein containing the C-terminal unique domain of Jessie3, which was purified with the amylose resin (left lane), fails to bind to chitin (right lane). Right center panel: A Western blot with polyclonal rabbit antibodies to Jessie3 shows that an MBP-fusion-protein containing full-length Jessie3 self-aggregates, so that it is difficult to purify on the amylose resin (left lane). However, the MBP-full-length Jessie3 fusion-protein binds to chitin (right lane). Right panel: Polyclonal rabbit anti-Jessie3 antibodies bind weakly to trophozoites of Ei (left lane) but bind strongly to an ∼60-kDa protein in encysting Ei (the expected size of Jessie3) (right lane). Lower molecular weight bands may reflect a cleavage product between the N-terminal CBD and the C-terminal unique domain, as we have previously shown that Jacob lectins of encysting Ei are often cleaved between CBDs . There was no binding of a control non-immune sera to trophozoite or cyst proteins (not shown).

Figure 6. Recombinant Jessie 3 lectins self-aggregate.

Fluorescence microscopy (A to D) and negative staining (E) of bacteria transformed with MBP-fusion-proteins expressing full-length Jessie3 (green in A) or the unique C-terminal unique domain of Jessie3 (green in B) agglutinate to form large clumps of bacteria. In addition, biofilms containing MBP-full length Jessie3 are formed (green in D and immunostained with gold in E). Antibodies to the unique C-terminal domain of Jessie 3 are conjugated to Alexafluor in (A, B, and D) or are detected with immunogold in (E). Antibodies to the N-terminal CBD of Jessie3 (green in C) show that MBP-fusion-proteins containing this domain do not self-aggregate on a macro-scale and do not agglutinate bacteria. Bar (A to D) is 2 microns. Bar (E) is 500 nm.

Figure 7. Negative stains of bacteria transformed with MBP constructs targeted to the periplasm suggest the unique C-terminal domain of Jessie3 is an important contributor to self-aggregation.

Bacteria (B) expressing MBP alone (low magnification in A and high magnification in F) or an MBP-Jacob2 fusion-protein (jac in low magnification in B) release dense quantities of proteins, which do not self-aggregate. In contrast, bacteria expressing an MBP-Jessie3 CBD fusion-protein (J3 CBD in low magnification in C and high magnification in G) release proteins in conical shaped eruptions from the periplasm, and the MBP-Jessie3 CBD fusion-proteins form short, thin linear arrays. Bacteria expressing MBP-Jessie3 unique C-terminal domain fusion-proteins (J3 sad (self-adhering domain) in low magnification in D and high magnification in H) or MBP-full-length Jessie3 (J3 ful in low magnification in E and high magnification in I) release dense aggregates of proteins, and sheets of these MBP-Jessie3 fusion-proteins contain branched aggregates of multiple-stranded linear forms that are thicker than those formed by Jessie3 CBD. Bar (A to E) is 250 nm. Bar (F to I) is 100 nm. Circles and ovals are added to figures to highlight structures of secreted proteins.

Figure 8. “Wattle and daub” model of the Ei cyst wall (a hypothesis).

In the first “foundation” phase of encystation, Jacob lectins, which are themselves glycoproteins that contain Gal, are bound to the surface of encysting amebae by constitutively expressed plasma membrane Gal/GalNAc lectins ,. In the second “wattle” phase, Jacob lectins, which contain tandemly arranged CBDs, appear to cross-link chitin fibrils that are deposited on the surface of encysting amebae. In the third “daub” phase, the cyst wall is solidified and made impermeable to small molecules by the addition of the Jessie3 lectin, which has an N-terminal CBD that binds chitin and a unique C-terminal unique domain that appears to cause self-aggregation.