Murine Raw264.7 macrophages (ATCC) were grown in DMEM F-12 (Cytiva, Marlborough, MA) supplemented with 1% penicillin/streptomycin and 5% fetal bovine serum (Cytiva, Marlborough, MA). Immunodeficient NU/NU mice were used in the amyloidoma experiments described hereafter. All animal studies described herein were carried out in accordance with protocols approved by the University of Tennessee Institutional Animal Care and Use Committee and in accordance with the guidelines provided by the Office of Laboratory Animal Welfare (OLAW) and the Guide for the Care and Use of Laboratory Animals. The University of Tennessee Graduate School of Medicine is an Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC)-accredited institution.

Amyloid-like fibrils composed of rVλ6Wil were prepared in PBS with 0.05% sodium azide as previously described (24). Briefly, λ6 WIL light chain variable domain monomers were purified from e. coli periplasmic fractions. Following lyophilization, protein was resuspended in PBS and shaken at 37°C to induce fibril formation. Fibril content was confirmed by Thioflavin T (ThT) fluorescence analysis and classical fibrillar conformation was confirmed by transmission electron microscopy. Human amyloid extracts were prepared from autopsy-derived tissues from patients with ALλ (designated CLA) or ALκ (TAL) associated amyloidosis using the water flotation method, as previously described (25). The use of human-subject–derived materials was approved by the University of Tennessee Graduate School of Medicine Institutional Review Board.

Human blood was drawn from healthy, consented donors under an IRB-approved protocol (University of Tennessee Graduate School of Medicine IRB no.: 4011). Neutrophils were purified from 20 mL of peripheral blood as previously described (26). Briefly, blood was diluted 2:1 in PBS and 10 mL lymphocyte separation media (Lonza Biosciences, Walkersville, MD) was underlaid. Centrifugation was used to separate the blood into peripheral blood mononuclear cells and a neutrophil and red blood cell pellet. Red blood cells were lysed using a hypotonic lysis, yielding isolated neutrophils. Neutrophil purity (based on CD66b-FITC expression) was confirmed by flow cytometry (Northern Lights Cytometer, Cytek, Fremont, CA), and was greater than 95%. Murine neutrophils were obtained via intraperitoneal injection of 3% Thioglycolate (27, 28). Whole peritoneal exudate was stained with CD11b and Ly6G to identify neutrophils which phagocytosed amyloid.

A suspension of patient-derived amyloid extract (2 mg extract in 0.2 mL), containing 10% (w/w) DyLight 800-labeled (DL800) extract, was injected subcutaneously, on the dorsal flank of male and female NU/NU mice (n = 8-12). Mice, under isoflurane anesthesia, were imaged serially, over a 15-day period after injection using an iBox Scientia small animal optical imaging system (Analytik Jena, Upland, CA), using an 800-nm bandpass filter set for DL800 (2-second exposures with 1 × 1 binning). Animals were euthanized by isoflurane overdose at day 18 post injection of amyloid material, and the residual masses were harvested and fixed in 10% buffered formalin (24 h) and paraffin embedded for histologic examination.

The mean raw density (MRD) of the amyloid-associated fluorescence in the DL800 images was performed by region of interest (ROI) analysis using VisionWorks v. 8.20 (Analytik Jena, Upland, CA). Freeform ROIs encompassing the amyloid mass were manually drawn on the images by a single reviewer who was blinded to the study design. A single ROI was used for each timepoint for each individual animal in the cohort. A second ROI covering a representative amyloid-free region of the mouse was drawn and served as the mouse-specific background fluorescence (associated with autofluorescence). The final fluorescence signal intensity was calculated by subtraction of the background mean raw density from the amyloid-associated fluorescence MRD.

Amyloidomas were harvested 18 days post implantation. Amyloidoma area was calculated as previously described (29) using NIH ImageJ software (http://rsbweb.nih.govij). Briefly, amyloidomas were imaged with a ruler in frame for scale. ImageJ software was used to convert 1 mm into pixel number, and subsequently, area of each amyloidoma was calculated.

Sections (6 μm thick) of residual amyloid embedded in paraffin were prepared on Plus slides and stained with hematoxylin and eosin (Fisher Scientific, Waltham, MA) or alkaline Congo red solution (0.8% w/v Congo red, 0.2% w/v KOH, and 80% ethanol) for 1 h at room temperature, followed by counterstain with Mayer’s hematoxylin for 2 minutes. In order to specifically identify intact mouse neutrophils in amyloidoma sections, we performed Ly6G clone 1A8 (Fisher Scientific, Waltham, MA) staining. Neutrophil extracellular traps were histologically identified by the co deposition of the DNA-reactive dye hematoxylin [blue], the presence of transmembrane neutrophil marker LY6G, and the presence of citrullinated histone 3, in the absence of intact whole neutrophils (30).

Photomicrographs were acquired using a BZ-X700E microscope (Keyence, Atlanta, GA). Congo red fluorescence images utilized a Texas red filter. Bright-field images and were typically acquired using a 40× objective and an exposure time of 1/40 seconds, unless otherwise noted.

Analysis was performed at the clinical mass spectrometry facility at the Mayo Clinic (Rochester). Briefly, nano-flow liquid chromatography electrospray tandem mass spectrometry was performed as previously described (31, 32). Peptide spectra present in the raw data files were matched against a composite protein sequence database using three different search engines (Sequest, X!Tandem, and Mascot).

Patient-derived amyloid material (20 μg) was placed onto a carbon coated microgrid and stained with UranyLess (Electron Microscopy Sciences, Hatfield, PA) solution for 1 minute. Samples were allowed to dry overnight. All the Transmission Electron Microscopy images were captured using the JEOL JEM 1400-FLAH TEM (JEOL USA, Inc. Peabody, Massachusetts, USA) at 120 kV with a Gatan OneView camera (Pleasanton, California, USA). Images were evaluated using ImageJ software.

In vivo depletion of neutrophils was performed using anti-Ly6G (clone 1A8) (Bio X Cell, Lebanon, NH) as previously described (28). Briefly, depleting antibodies were administered (0.25 mg) every day starting at 2 days prior to amyloid implantation and then every other day for 7 days. Flow cytometry and immunohistochemistry were used to confirm neutrophil depletion.

Human ALλ(CLA) or ALκ(TAL) extracts (2mg) were implanted subcutaneously into NU/NU mice. Amyloidomas were harvested 1 day and 8 days post implantation. Tissue was mechanically dissected and treated with type-IV collagenase for 30 minutes with shaking at 37°C. Digested samples were passed through a 70 μm filter to obtain single cell suspensions. Cells were then stained with the following antibodies: Live/Dead Aqua (Invitrogen, Thermo Fisher Scientific, Waltham, MA), CD45 (clone 30-F11), CD11b (clone M1/70), Ly6G (clone 1A8), Ly6C (clone HK1.4), F4/80 (clone BM8). All antibodies were obtained from BioLegend (San Diego, CA) unless otherwise specified. Following staining, samples were analyzed using a Cytek Northern Lights flow cytometer equipped with 405 nm and 488 nm lasers (Cytek, Fremont, CA). A Boolean gating strategy was used to remove signal from debris and dead cells, and to identify cellular populations. Data were analyzed using FlowJo software version 10.3.

For phagocytosis assays, between 1x10⁵ and 5x10⁵ primary human neutrophils or whole murine peritoneal exudate were incubated with 20 μg of patient-derived amyloid extract labeled with the pH-sensitive fluorophore pHrodo Red (20% w/w pHrodoTM red) for 1 h, 3 h, or 6 h at 37°C in a 5 mL (12mmx75mm) polystyrene tube, then analyzed for amyloid uptake using a 2-laser Northern Lights flow cytometer (Cytek, Fremont, CA). Peritoneal exudate was stained as above to identify neutrophils which phagocytosed amyloid. A Boolean gating strategy was used to identify live, single cells. Data were analyzed using FlowJo software version 10.3.

To parameterize neutrophil oxidative burst following exposure to amyloid extract, purified human neutrophils were assessed using a total ROS (reactive oxygen species) staining kit (Invitrogen, Thermo Fisher, Waltham, MA). ROS production was quantified using a Northern Lights flow cytometer (Cytek, Fremont, CA).

To analyze cytokine expression in the excised amyloidomas following neutrophil-depletion, extracted amyloidoma tissue was mechanically homogenized in 100 μL PBS and supernatants processed using the BioLegend, LEGENDplex™ mouse anti-virus response panel. Data were collected using a Northern Lights flow cytometer (Cytek, Fremont, CA). Data was analyzed using the LEGENDplex™ Data Analysis Software Suite from Qognit.

An aliquot of 100 μg of patient-derived amyloid extract or purified rVλ6WIL fibrils were incubated overnight in a 30% H₂O₂ 70% PBS solution (non-chelex treated), or PBS alone. Following incubation, amyloid was washed twice via centrifugation at 1000 x g and resuspended in PBS. Fibril content post-H2O2 treatment (or PBS control) was analyzed by Thioflavin T fluorescence emission.

For direct comparison of two normally-distributed variables, unpaired, two-tailed Student’s t-tests (α = 0.05) were used. Comparison of mean raw density fluorescence data between initial, peak, and subsequent time points, when normally distributed, was performed by using a mixed effect ANOVA with Tukey’s multiple comparison (α = 0.05). All statistical analyses were performed using Prism software version 9.01 (GraphPad Software Inc., San Diego, CA).

All patient-derived tissue samples were used in accordance with an Institutional Review Board-approved application. Animal studies were approved by the University of Tennessee Institutional Animal Care and Use Committee and were performed in accordance with the guidelines provided by OLAW and the Guide for the Care and Use of Laboratory Animals. The University of Tennessee Medical Center animal program is AAALAC-i-accredited.

Both human AL extracts were composed of sheets of linearly aligned, unbranching fibrils, when imaged using transmission electron microscopy. In addition, single, non-twisting, fibrillar structures were present in both samples. Both ALλ(CLA) and ALκ(TAL) appeared ultrastructurally comparable ( Figures 1A, B ). Addition of ThT resulted in similar fluorescence emission at 490 nm, characteristic of the presence of amyloid fibrils, although ALκ(TAL) exhibited significantly greater fluorescence emission suggesting the presence of either more amyloid fibrils per unit mass of extract, more ThT binding sites due to differences in fibril morphology, or higher order aggregation states Figure 1C.

Mass spectrometric analysis of the amyloid extracts demonstrated near-identical composition of the materials with respect to t known amyloid-associated accessory molecules (except for the presence of histone H2A in the ALκ preparation). Differences in the number of recovered peptides, potentially reflects discordance in their relative abundance in the extracts ( Table 1 ).

The clearance of both AL amyloid extracts implanted subcutaneously in NU/NU mice, based on the decrease in DL800 fluorescence emission, was shown to be significantly slower for ALλ(CLA) as compared to ALκ(TAL) amyloid extract ( Figure 2 ). Relative to the fluorescence emission at Day 1, by Day 15 post-injection the fluorescence emission associated with the ALλ(CLA) amyloidoma had not decreased, whereas that from ALκ(TAL) was ~75% reduced.

Since ALλ(CLA) and ALκ(TAL) amyloid clearance rates differed in vivo, we sought to determine whether immunological changes within the two patient-derived amyloidomas could be observed. Amyloidomas were harvested from NU/NU mice 1 day and 8 days post implantation and immunophenotyping was performed to identify immune cell infiltrates (See Supplementary Figure 1 for the gating strategy used for analysis). Phenotypically, 1 day post amyloid implantation there was no difference in the number of live myeloid cells (CD45^+,CD11b⁺) isolated from either the ALλ(CLA) and ALκ(TAL) amyloidomas. However, the composition of the myeloid cell compartment varied as the total number of macrophages (CD45^+,CD11b⁺, Ly6C⁺, Ly6G^-, F4/80⁺) identified in ALλ(CLA) amyloid was significantly greater than that in seen in ALκ(TAL) amyloid. There was no difference in the number of live neutrophils (CD45^+,CD11b⁺, Ly6C⁺, Ly6G⁺, F4/80^-) or monocytes (CD45^+,CD11b⁺, Ly6C⁺, Ly6G^-, F4/80^-) identified between ALκ and ALλ amyloidomas at 1 day post amyloid implantation. Eight days post amyloid implantation, ALλ(CLA) amyloidomas contained significantly more live myeloid cells as well as a continued significant increase in macrophages compared to ALκ(TAL). However, ALκ(TAL) exhibited significantly more neutrophils compared to ALλ(CLA) ( Figure 3B ). There was no difference in monocyte content 8 days post amyloid implantation. Finally, overall myeloid cell composition between ALλ(CLA) and ALκ(TAL) amyloidomas was drastically different ( Figures 3A, B , far right panels %CD11b).

To spatially interrogate intra-amyloid immune cell distribution, tissue sections were analyzed 1 day and 8 days post ALλ(CLA) and ALκ(TAL) amyloidoma implantation. Staining of the amyloidomas excised at day 1 post injection with H&E showed pronounced infiltration of polymorphonuclear cells (i.e., neutrophils) throughout the ALκ(TAL) amyloidoma with profound DNA deposition ( Figure 3D black arrows). There was no observable DNA deposition observed in ALλ(CLA) amyloidomas and the presence of neutrophils was markedly reduced relative to the ALκ(TAL) amyloidoma ( Figure 3C black arrows. Subsequent evaluation of amyloidomas harvested 8 days post implantation indicated that ALλ(CLA) amyloid did have immune infiltrate; however, ALκ material again contained identifiable polymorphonuclear cells, large macrophage-like cells, and multinucleated giant cells. Next, we subjected ALλ(CLA) and ALκ(TAL) amyloidomas harvested 8 days post implantation to a combinatorial staining regime to specially analyze neutrophils by combining H&E staining, the neutrophil marker, Ly6G, and the neutrophil extracellular trap marker, citrullinated histone 3. ALλ(CLA) amyloidomas were devoid of neutrophils 8 days post amyloid implantation as evidence by both H&E staining and Ly6G staining ( Figure 4A , left and middle panels). In contrast, ALκ(TAL) amyloidomas contained a plethora of Ly6G staining including both intact neutrophils and diffuse intra-amyloidoma Ly6G staining, suggesting neutrophil death ( Figure 4B ). The diffuse intra-amyloidoma Ly6G staining co-localized with pronounced citrullinated histone 3 indicating the presence of neutrophil extracellular traps. Importantly, co-citrullinated histone 3 and Ly6G staining was not observed in the intact neutrophil population, ( Figure 4B right panel).

Based on the amount of observable NETosis and abundance of identifiable cellular/nuclear debris in the ALκ amyloidomas, we reexamined the flow cytometry data in Figures 3A, B this time including non-viable cells which were identified using an amine reactive viability dye (traditionally removed using a standard Boolean gating strategy Supplementary Figure 1 ). The proportion of viable CD45⁺ immune cells compared to the total population of immune cells was 2.5-fold (p ≤ 0.0001) lower in ALκ(TAL) 1 day following amyloid implantation and at 8 days post implantation of the amyloid ALκ(TAL) amyloidomas contained 3.75-fold (p ≤ 0.0001) fewer viable immune cells compared to ALλ(CLA) ( Supplementary Figure 2 ). Inclusion of non-viable cells in our analysis identified significantly more neutrophils at both day 1 (p=0.0044, 2.25-fold) and day 8 (p=0.0022, 9.75-fold) post amyloid implantation in ALκ(TAL) compared to ALλ(CLA) ( Supplementary Figure 2 ). Examination of the remaining immune cells was not significant when including non-viable cells, but the number of identified monocytes in ALκ(TAL) amyloidomas trended towards an increase and the number macrophages isolated from either amyloidoma was virtually identical ( Supplementary Figure 2 ).

With the stark differences in immune cell infiltrates between ALλ(CLA) and ALκ(TAL) amyloidomas we sought to evaluate their immunological activation status. Using the BioLegend, LEGENDplex™ cytokine expression platform we analyzed cytokine expression at 8 days post amyloid implantation. ALλ(CLA) amyloidomas contained significantly more CXCL10, GM-CSF (granulocyte monocyte-colony stimulating factor), and IL-10, with more expression of CCL5 trending towards significance (p=0.0882) ( Figure 5 ). ALκ(TAL) amyloidomas had significantly more expression of monocyte chemoattractant protein-1 (MCP1; CCL2), and more expression of tumor necrosis factor-alpha trending towards significance (p=0.0627) ( Figure 5 ).

We analyzed neutrophil activity ex vivo following exposure to amyloid extract and amyloid-like synthetic fibrils. To assess the ability of neutrophils to directly phagocytose ALλ(CLA) extract, ALκ(TAL) extract, or purified synthetic rVλ6WIL fibrils, pHrodoRed-labeled substrates were exposed to either human or mouse neutrophils in vitro and analyzed by flow cytometry. In this ex vivo system, purified human neutrophils would not or could not phagocytose any amyloid material ( Figure 6A ). Also murine neutrophils from whole thioglycolate-elicited peritoneal exudate did not phagocytose amyloid extracts and amyloid-like fibril phagocytosis was minimal ( Figure 6B ). We next evaluated the reactive oxygen burst of human neutrophils following exposure to the amyloid substrates and, as a positive control, to formylated-met-leu-phe peptide (fMLF). Neither amyloid extract induced an oxidative burst in vitro ( Figure 6C ). However, since in vivo amyloidoma conditions are more complex than in vitro conditions, we questioned whether, if generated in vivo, what effect reactive oxygen species (e.g., H₂O₂) would have on the amyloid extracts and synthetic fibrils. Hydrogen peroxide treatment of both rVλ6Wil fibrils and ALκ(TAL) extract significantly reduced the ThT fluorescence emission associated with the substrates. In contrast, ALλ(CLA) amyloid was refractory to H₂O₂ treatment, and no significant change in ThT fluorescence emission was observed ( Figure 6D ).

Given the neutrophil influx and resultant NETosis in the ALκ(TAL) amyloidoma and their potential impact on the amyloid mass, we assessed the impact of neutrophil depletion on ALκ(TAL) dissolution in vivo ( Figure 7 ). Neutrophil depletion, accomplished using the anti-Ly6G (1A8) antibody and confirmed by flow cytometry and histological analysis ( Supplementary Figure 3 ), significantly delayed ALκ(TAL) amyloidoma clearance in NU/NU mice as evidenced by changes in DL800 fluorescence intensity over 18 days post amyloid injection ( Figures 7A, B ). The amyloid mass at days 11, 13, and 15 post-injection was significantly less (as measured via fluorescence) in untreated mice as compared to neutrophil-depleted animals ( Figures 7B, D ). Furthermore, measurement of the area of the residual amyloid masses, harvested from the mice at necropsy, showed ~2-fold larger amyloid lesions in mice treated with neutrophil-depleting antibodies ( Figure 7C ).

To further interrogate the mechanisms by which neutrophils may impact ALκ(TAL) clearance, we monitored the immune activation status of the ALκ(TAL) amyloid excised from neutrophil-depleted and naive mice. Cytokine and chemokine expression was measured 1- and 4-days post implantation by multiparameter flow cytometric ELISA-based assays. Day 1 and day 4 were chosen as both were within the active neutrophil depletion window ( Figure 8A ) before depletion pressure was removed. Amyloid from neutrophil-depleted mice had significantly more CXCL1, a neutrophil chemokine, compared to amyloid from non-depleted animals, which had significantly more expression of MCP-1. No significant differences were observed in amyloid associated CCL5 (RANTES), IL-1β, or GM-CSF. At day 4, naïve animals had significantly more expression of MCP-1, CCL5, as well as the inflammatory mediators IL-1β and GM-CSF ( Figure 8B ) compared to amyloid excised from neutrophil-depleted mice.

Histological evaluation of ALκ(TAL) amyloid resected from neutrophil-depleted NU/NU mice or non-treated NU/NU mice on days 8, 11, and 18 following depletions indicate a delay in neutrophil infiltration of the amyloidoma in mice depleted of neutrophils compared to untreated-mice. Non-depleted mice had a concomitant accumulation of multinucleated giant cells as well as polymorphonuclear cells within the amyloid deposits at earlier time points (Day 8) compared to neutrophil-depleted mice (Day 11) ( Figure 9 ). Furthermore, the immunological landscape and amyloidoma structure of neutrophil-depleted day 8 ALκ(TAL) amyloidomas and day 8 ALλ(CLA) amyloidomas ( Figures 3C, D ) were visually similar.