Targeting cargo to an unconventional secretory system within megakaryocytes allows the release of transgenic proteins from platelets

Abstract

Background: Platelets are essential for hemostasis and thrombosis and play vital roles during metastatic cancer progression and infection. Hallmarks of platelet function are activation, cytoskeletal rearrangements, and the degranulation of their cellular contents upon stimulation. While α-granules and dense granules are the most studied platelet secretory granules, the dense tubular system (DTS) also functions as a secretory system for vascular thiol isomerases. However, how DTS cargo is packaged and transported from megakaryocytes (MKs) to platelets is poorly understood.

Objectives: To underpin the mechanisms responsible for DTS cargo transport and leverage those for therapeutic protein packaging into platelets.

Methods: A retroviral expression system combined with immunofluorescence confocal microscopy was employed to track protein DTS cargo protein disulfide isomerase fused to enhanced green fluorescent protein (eGFP-PDI) during platelet production. Murine bone marrow transplantation models were used to determine the release of therapeutic proteins from platelets.

Results: We demonstrated that the endoplasmic reticulum retrieval motif Lys-Asp-Glu-Leu (KDEL) located at the C-terminus of protein disulfide isomerase was essential for the regular transport of eGFP-PDI-containing granules. eGFP-PDI^ΔKDEL, in which the retrieval signal was deleted, was aberrantly packaged, and its expression was upregulated within clathrin-coated endosomes. Finally, we found that ectopic transgenic proteins, such as tissue factor pathway inhibitor and interleukin 2, can be packaged into MKs and proplatelets by adding a KDEL retrieval sequence.

Conclusion: Our data corroborate the DTS as a noncanonical secretory system in platelets and demonstrate that in vitro-generated MKs and platelets may be used as a delivery system for transgenic proteins during cellular therapy.

Keywords: granules; hemostasis; megakaryocyte; platelet; protein disulfide isomerase.

Figure 1:. PDI is localized to distinct organelles and distributed differently than canonical α-granule and δ-granule markers.

(A-D) Immunofluorescence imaging of mature fetal-liver MKs, proplatelets, and platelets labeled with α-tubulin, DAPI, and either of the following: (A) vWF (α-granules), (B) Rab27b (δ-granules), (C) Calnexin (ER), or (D) PDI (ER). Scale bars represent 10 μm in MK imaging or 5 μm in platelet imaging. (E) Electron microscopy of ultrathin frozen sections of resting human platelets labeled with primary PDI antibody followed by secondary attached to immunogold particles of 15 nm diameter. Cyan arrows indicate immunogold labeling of PDI, and green arrows indicate α-granules. (F) Electron microscopy of ultrathin cryosections of activated human platelets labeled with PDI antibody, cyan arrows indicate PDI labeling, green arrows indicate α-granules and magenta arrows indicate PDI release at the surface of activated platelets. Scale bars represent 100 nm.

Figure 2:. Overexpressed eGFP-PDI does not colocalize with α-granules.

(A) Plasmid map of eGFP-PDI MSCV puro. (B) Linear schematic of eGFP-PDI fusion protein insert. Human LDL ER signal peptide (blue), enhanced green fluorescent protein (eGFP) green, followed by TGGG linker (black) and mature mouse PDI protein (magenta). (C) Schematic of fetal liver transduction protocol. Image was created using Biorender.com. (D) Western blots for PDI in fetal-liver MK cell lysates. Left and right blot are technical repeats of: Lane 1: lysate of non-transduced fetal-liver MKs, lane 2: lysate from eGFP-PDI transduced fetal liver MKs. Lane 3: lysate from eGFP-PDI^ΔKDEL transduced fetal liver MKs. (E) Co-localization analysis (Manders coefficient) of eGFP-PDI with PDI or vWF. Colocalization measurements were performed on round megakaryocytes (●) and proplatelet-forming megakaryocytes (△). (F-G) Immunofluorescence micrographs of eGFP-PDI transduced mouse fetal liver MKs and proplatelets white = CD41⁺, cyan = eGFP-PDI, magenta = either PDI or vWF. Scale bars represent 10 μm.

Figure 3:. Removal of ER retrieval motif leads to delayed and disrupted packaging of PDI protein throughout proplatelets.

(A-B) Electron microscopy of ultrathin frozen sections of mouse fetal liver megakaryocytes (A) and mouse platelets (B) labeled with primary PDI antibody and secondary immunogold particles of 15 nm diameter. Cyan arrows indicate immunogold labeling of PDI, and green arrows indicate multi-vesicular bodies (in megakaryocytes) and α-granules in platelets. (C-D) Fetal liver megakaryocytes transduced with either eGFP-PDI (C) or eGFP-PDI^ΔKDEL. The white dotted line indicates the megakaryocyte and proplatelet outline. (E) Rate of movement of PDI-containing particles. (F) Total distance traveled of PDI- PDI-containing particles. Statistical significance was determined using pared unequal variance tests.

Figure 4:. Removal of KDEL retrieval motif leads to abnormal distribution and increased packaging of eGFP-PDI into clathrin-coated vesicles.

(A-E) Immunofluorescence micrographs of transduced fetal liver MKs and proplatelets with either eGFP-PDI or eGFP-PDI^ΔKDEL. Blue = DAPI, white = CD41⁺, cyan = eGFP-PDI, magenta = (A) LAMP-1, (B) Calnexin, (C) COPI, (D) COPII, (E) Clathrin. Scale bars represent 10 μm. (F) Co-localization analysis (Manders coefficient) of eGFP-PDI or eGFP-PDI^ΔKDEL with calnexin, COPI, COPII, clathrin, and LAMP-1 antibodies. Statistical significance was determined using ordinary one-way ANOVA with multiple comparison tests.

Figure 5:. Transgenic proteins can be packaged into hematopoietic stem and progenitor cells, resulting in mature murine circulating transgenic platelets that can release cargo upon activation.

(A-D) Immunofluorescence microscopy of eGFP-PDI, eGFP-PDI^ΔKDEL, eGFP-TFPI and eGFP-IL-2 co-stained with ER biomarker calnexin. Scale bars represent 10 μm. (E) Co-localization analysis (Manders coefficient) of eGFP-PDI, eGFP-PDI^ΔKDEL, eGFP-TFPI, and eGFP-IL-2 with calnexin. (F) Frequency (%) of donor or host cells in hemolysed blood isolated from bone marrow recipients three weeks post-transplantation. (G) Percentage (%) of eGFP⁺ cells throughout other blood lineages three weeks post-transplantation. (H) Percentage of eGFP⁺ platelets isolated from mouse blood collected from bone marrow transplanted mice or untransplanted controls. (I) eGFP MFI of resting and activated platelets. (J) Integrin activation upon stimulation of platelets transduced with either eGFP control, eGFP-IL-2, or eGFP-TFPI with 0.1 U/mL thrombin, data points analysed using 2-way ANOVA with dunnets multiple comparison. (K) Quantity of eGFP present within resting or activated platelet releasate (pg/mL) for either eGFP control, eGFP-IL-2, or eGFP-TFPI. Data points within (E-J) represent individual biological samples from five independent mice, data points. were analysed using multiple unpaired t-tests.