GlRac regulates a noncanonical, ATG8-independent autophagy-like pathway in Giardia lamblia
Data files
Sep 28, 2026 version files 165.61 MB
-
Fig1.zip
112.59 KB
-
Fig2.zip
328.02 KB
-
Fig3.zip
169.39 KB
-
Fig4.zip
118.32 MB
-
Fig5.zip
558.54 KB
-
Fig6.zip
393.74 KB
-
Fig7.zip
260.68 KB
-
Fig8.zip
209.97 KB
-
FigS10.zip
1.14 MB
-
FigS11.zip
40.79 MB
-
FigS12.zip
70.57 KB
-
FigS3.zip
37.18 KB
-
FigS4.zip
36.88 KB
-
FigS6.zip
2.41 MB
-
FigS7.zip
748.46 KB
-
README.md
21.43 KB
Abstract
Autophagy is a conserved catabolic process essential for cellular homeostasis and stress adaptation. The protozoan parasite Giardia lamblia lacks most canonical autophagy-related (ATG) genes, including the hallmark ATG8, raising longstanding questions about whether it can perform autophagy. Here, we show that Giardia mounts a regulated autophagic-like response. Double-membrane compartments resembling autophagosomes are induced in up to 30% of encysting cells and 91% of starved trophozoites, supporting roles in differentiation and survival under nutrient stress. Their clearance is triggered by amino acid replenishment but not by glucose, indicating a nutrient-specific sensing mechanism. GlRac, the parasite’s sole Rho family GTPase, labels these structures and regulates their formation, as evidenced by a threefold increase in compartment levels upon constitutive activation and a significant reduction after knockdown. This extends the conserved role of Rho GTPases in regulating autophagy to an evolutionarily early-branching eukaryote. Of nine putative ATG orthologs tested, none localized as clearly as GlRac to autophagic structures. These organelles acidify and recruit cathepsin proteases, consistent with degradative capacity. A newly developed live-cell actin marker reveals robust recruitment to these structures, implicating actin-driven remodeling. Finally, quinacrine, an FDA-approved antigiardial drug, promotes the accumulation of autophagic structures, consistent with its known effects on mammalian autophagy. Together, our findings establish GlRac as a regulator of an ATG8-independent autophagic response in Giardia, demonstrate that this parasite retains key features of autophagy despite its streamlined genome, and highlight this pathway as a potential therapeutic target.
This dataset contains the source data underlying the figures of the study "GlRac
regulates an ATG8-independent autophagy-like pathway in Giardia lamblia." The study
characterizes GlRac-positive double-membrane compartments in the protozoan parasite
Giardia lamblia as a divergent, ATG8-independent autophagy-like pathway, using live-cell
fluorescence microscopy, time-lapse imaging, bioluminescence assays, morpholino knockdown,
constitutive activation, and pharmacological perturbation.
Related publication
Klippel AH, Newman-Boulle C, Reed G, Dule D, Kedia S, Kelty MT, Chan WP, Shih H-W,
Paredez AR. GlRac regulates an ATG8-independent autophagy-like pathway in
Giardia lamblia. [Journal — add citation and DOI on acceptance.]
- Dryad dataset DOI: https://doi.org/10.5061/dryad.bzkh189sc
- Corresponding author: Alexander R. Paredez (aparedez@uw.edu), Department of Biology,
University of Washington, Seattle, WA, USA. - Funding: National Institutes of Health grants R01AI168417 and R21AI159035 (to A.R.P.).
Description of the data and file structure
The data are organized by manuscript figure. Each figure is uploaded as a single ZIP
archive (e.g. Fig1.zip) that unpacks to a correspondingly named folder (e.g. Fig1/)
containing the source data for that figure's quantified panels: the raw per-replicate
measurements (as plain .csv tables and as the original Excel workbooks, .xlsx or .xls,
in which they were recorded; see "Tabular data: CSV files and original Excel workbooks" below), the
corresponding GraphPad Prism analysis/graph files (.prism),
and, for figures reporting localization or morphology, representative deconvolved
micrographs (.pdf) or time-lapse videos (.avi, .tif, .gif). Filenames encode the
panel, the acquisition date (YYYYMMDD or MMDD), the cell line/treatment, and the
replicate.
Files in this deposit:
README.md This file
Fig1.zip -> Fig1/ GlRac localizes to a double-membrane organelle induced by encystation
Fig2.zip -> Fig2/ Nutrient availability vs. GlRac-positive compartments (aging, refeeding, individual nutrients)
Fig3.zip -> Fig3/ GTOR levels (NanoLuc) and GTOR morpholino knockdown
Fig4.zip -> Fig4/ Time-lapse videos of compartment biogenesis and clearance
Fig5.zip -> Fig5/ Conserved features: concanamycin A, acidification (LysoTracker), cathepsin coincidence
Fig6.zip -> Fig6/ Active cysteine proteases (BODIPY-LHVS) and E-64d treatment
Fig7.zip -> Fig7/ GlRac knockdown and constitutive activation (CA-Rac)
Fig8.zip -> Fig8/ Quinacrine treatment
FigS3.zip -> FigS3/ Rapamycin treatment
FigS4.zip -> FigS4/ Anti-GTOR morpholino NanoLuc knockdown validation
FigS6.zip -> FigS6/ Localization of GlATG candidates (mNG)
FigS7.zip -> FigS7/ Proposed ATG7/ATG8 candidates (mNG)
FigS10.zip -> FigS10/ Additional cathepsins (mNG)
FigS11.zip -> FigS11/ Anti-GlRac morpholino immunoblot (densitometry + membrane scan)
FigS12.zip -> FigS12/ GlRac knockdown penetrance
Note on scope: figure and panel numbering follows version 2 of the preprint
(bioRxiv, doi:10.64898/2026.02.17.706457v2). This deposit holds the quantitative source data
for the quantified panels, the Fig 4 time-lapse source movies, and selected representative
micrographs (Fig5H, Fig6A-B, S6, S7, and S10 Figs). Other representative images (for
example Fig 1A and 1C, the TEM panels in S2 Fig, and S1, S5, S8, and S9 Figs) and the S1 and
S2 Datasets are provided with the article rather than here.
Tabular data: CSV files and original Excel workbooks
All tabular data are provided in two formats, side by side in each figure folder:
- CSV files (
.csv), one per worksheet. We added these at the request of Dryad
curation. They are plain, comma-delimited text with no formatting, formulas, merged
cells, or highlighting, and can be opened with any text editor, spreadsheet program, or
statistical software. For reanalysis, use the CSV files. - The original Excel workbooks (
.xlsx, and one.xls) in which the data were recorded during
scoring. We have kept these on purpose alongside the CSVs. Each workbook groups all
of an experiment's conditions or timepoints as separate sheets in one file. Its
header layout, cell merges, and summary formulas show how the per-image counts add
up to the totals and percentages plotted in the figures. We think seeing the data in
the layout it was recorded in adds context and makes the dataset easier to follow.
The workbooks open in Microsoft Excel or free alternatives such as LibreOffice Calc.
How the two formats correspond. The deposit contains 56 workbooks (55 .xlsx and one .xls) and 150 CSV files,
and every worksheet has exactly one CSV. The CSV is named after its workbook and sheet:
- Multi-sheet workbooks:
<workbook name>__<sheet name>.csv(two underscores; spaces and
other special characters in the sheet name replaced by_). Example: sheet0 hof
Fig1B_20240214_Halo_Rac_encystation_0-4h_rep1.xlsxis
Fig1B_20240214_Halo_Rac_encystation_0-4h_rep1__0_h.csv. - Single-sheet workbooks: the same base name with a
.csvextension. Example:
Fig1D_Compartment_sizing_encystation.xlsandFig1D_Compartment_sizing_encystation.csv.
The CSVs keep each sheet's row and column layout, so cell positions line up with the
workbook. Where a workbook cell contains a formula, the CSV holds the calculated value.
The CSVs are UTF-8 encoded; dates are written as YYYY-MM-DD; cells that contain only a
space in the workbook are left empty; and non-breaking spaces in text are written as
ordinary spaces. We checked every CSV cell by cell against its worksheet, and the data
values are identical in both formats.
Formatting in the workbooks and what it means. None of the formatting holds
information that is missing from the CSVs:
- Formulas only derive summary values from the raw counts in the same sheet:
column totals (SUMrow below the per-image counts); percent of cells with a
GlRac-positive compartment, calculated as (Cup + Hollow + Full) / Total cells × 100;
NanoLuc luminescence averaged across technical wells and expressed as percent of
control (Fig3C, FigS4; FigS4 also gives the standard deviation);
Halo/Tubulin ratios and percent reduction (FigS11); and percent signal coincidence (Fig5I). All of these appear in the
CSVs as calculated values. - Merged cells are used for titles and group headers, such as experiment name,
Replicate 1/2/3, orCells with compartmentsspanning the Cup/Hollow/Full columns.
In the compartments-per-cell sheets (Fig2C, Fig3E, Fig5E, Fig6D, Fig7C, Fig7G, Fig8E),
the Pic ID is also merged down across every cell scored in that image. In the CSV, the
text of a merged range appears only in its first (top-left) cell. A blank Pic ID in
a compartments-per-cell CSV therefore means the row belongs to the image named in the
nearest non-blank Pic ID above it. - Highlighting: yellow fill on the
Cup/Hollow/Fullheaders and sheet titles, and
alternating shading of values and Pic IDs in the compartment-sizing sheets, was a
visual aid during scoring. It does not encode groups, exclusions, or any other data
and can be ignored. - Instrument export sheets (
Result sheetin Fig3C and FigS4) keep the
colors and layout of the Tecan Spark plate-reader export. The colors come from the
instrument software and carry no additional information. - The workbooks contain no comments, hyperlinks, embedded figures, frozen panes, or
filters. The only hidden element is one empty column (column C) in
FigS3_20241127_Rapamycin_autophagy_quantification_Rep2.xlsx.
Cells outside the main data table. Because each CSV keeps its sheet's layout, some
CSVs contain summary cells or small blocks below or beside the main table. These are
listed below by file type (row and column positions are the same in the CSV and the
workbook; spreadsheet cell references are given, e.g. B8 = column 2, row 8).
- Proportion scoring sheets: one row per image, with the number of cells scored and the
number of cells whose GlRac-positive structure was cup-shaped, hollow, or full. Two
column layouts are used. In Fig1B, Fig2B, Fig2F, Fig2G, Fig7B, Fig7F, and FigS12 the
columns are Pic ID, Total cells, Cup, Hollow, Full; in Fig3D, Fig5D, Fig6C, and Fig8D
they are ID, Cup, Hollow, Full, Total. The totals row below the per-image rows
(labelledSUMin some files, unlabelled in others) holds the column totals, and the
single value in the next column, under the%label, is the percent of scored cells
with a GlRac-positive structure, (Cup + Hollow + Full) / Total × 100. This
per-replicate percentage is the value plotted in the figure. - Signal-coincidence sheets (Fig5I cathepsin and LysoTracker files): the totals row
gives column totals; the value under%is the percent of compartment-containing
cells whose compartments were marker-positive (Colocalize?total /Number of cells
total). In the cathepsin files,Total cellsis the sum of cells across the Cup,
Hollow, and Full stages. In
Fig5I_Actin_autophagosome_colocalization_scoring.csv, the rows labelledtotals
and% of cells with autophagosomes w/ Halo-Actin enrichmentgive the totals and
percentage for each of the two dated experiments (2024-07-31 and 2024-08-15). The
DALGreen values plotted in Fig 5I (100% in each of two experiments) are recorded only in
Fig5I_Colocalization_conserved_elements.prism. - Rapamycin sheets (
FigS3_2024*_Rapamycin_autophagy_quantification_Rep*.csv): each
sheet has two side-by-side blocks,A(columns E to H) andB(columns M to P), one per
condition, with atotalrow at the bottom. The condition order differs between the
two files: in Rep1 (20241121), A = DMSO control and B = 36 µM rapamycin; in Rep2
(20241127), A = 36 µM rapamycin and B = DMSO control (as labelled in cells D10 and L10).
The% cells with autophagosomescolumn shows#DIV/0!for unused image rows (image
numbers with no cells entered); these are empty rows, not missing data. The Prism file
(FigS3_Rapamycin_graph.prism) plots three independent experiments; the first two are
these workbooks (DMSO 17.0% vs rapamycin 19.3%, and DMSO 9.8% vs rapamycin 13.2%). - GTOR-NLuc 24 h vs 72 h (
Fig3C_*__Replicates_C2.csv): rows 2 to 7 (REP1to
REP6) are six NanoLuc measurements, three from each of two independent experiments
(REP1toREP3= experiment 1,REP4toREP6= experiment 2). Columns B and C are
raw luminescence (RLU) at 24 h and 72 h. CellB8is the mean 24 h RLU of all six
measurements; columns D and E (rel) express each measurement's RLU as a percent of
that mean. Columns D and E are the values plotted in Fig 3C. - GTOR morpholino NanoLuc (
FigS4/GTOR_KD_*__Cycle_3.csv): columns B to D are
technical-replicate RLU readings for three control and three GTOR KD samples, and
column E is their mean.F4(Average Ctrl) is the mean of the three control samples;
column G (Relative) expresses each sample as a percent ofF4.H4/I4and
H7/I7are the mean and standard deviation of the relative values for the control
and KD groups, andJ4:K5restates the two group means (Control = 100; GTOR KD = 30.1). - Immunoblot densitometry (
FigS11/Quantification.csv; the table starts in row 2):
columns B to E are two knockdown/control pairs,Rac KD1,Control1,Control2,
Rac KD2. RowsAnti-HaloandTubulinare background-subtracted band intensities
(arbitrary units), andRatiois Anti-Halo / Tubulin. The unlabelled row below it (row 7)
gives each ratio relative to its paired control (KD1 / Control1 in column B; KD2 /
Control2 in column E; each control = 1). The number in column B of the row reading
% reduction(row 9) is 100 × (1 − KD1 / Control1) = 70.3%, the reduction reported in
S11 Fig. Columns I and J (rows 2 to 9) list the same eight band intensities numbered 1
to 8, in the order Anti-Halo then Tubulin for KD1, Control1, Control2, and KD2; they are
the same numbers arranged in the table to their left.
File formats in this deposit
.csv— comma-delimited text tables, one per worksheet of the accompanying.xlsx
workbooks (see above). This is the recommended format for reanalysis..xlsx,.xls— the original workbooks in which per-cell and per-field counts and
measurements were recorded, one workbook per biological replicate or per figure panel.
They are kept alongside the CSVs for context (see above)..xlsis the older Excel
97-2003 format (used only for Fig1D). Openable in Excel, LibreOffice, or any
spreadsheet reader..prism— GraphPad Prism project files containing the compiled data, statistical
tests, and graphs for the panel. Proprietary to GraphPad Prism (v10). The underlying
numbers are also present in the accompanying.csvand Excel files, except where noted
for FigS3..pdf— representative deconvolved fluorescence micrographs used for the localization
panels (each named by the tagged gene, e.g.137680-mNG.pdf)..avi,.tif,.gif— time-lapse live-cell imaging clips (Fig4)..mscn— original Bio-Rad Image Lab scan of the FigS11 western blot membrane
(proprietary). Open 16-bit grayscale.tifexports of both channels are provided
alongside it..tif— 16-bit grayscale TIFF exports of the FigS11 blot channels (and a Fig4 clip).
Variables and conventions
Because the spreadsheets were recorded during scoring, several use side-by-side replicate
blocks and section headers rather than a single flat table. The recurring variables are:
- Pic ID — identifier of the source image/field (encodes date, cell line, and
timepoint/treatment). - Total cells — number of cells scored in that field.
- Compartment morphology categories —
Cup,Hollow, andFull, counting cells (or
compartments) at each morphological stage of the GlRac-positive compartment. "Cells with
compartment" is the sum across these categories. - Compartments per cell — integer count of GlRac-positive compartments in an
individual cell (Cell 1,Cell 2, …), grouped underReplicate 1/2/3(/4). - Compartment size — compartment diameter in micrometers (µm), one value per
compartment, grouped by replicate; sheet headers give the per-replicate n (e.g.
Rep 1 (n=35)). - NanoLuc luminescence (Fig3C, FigS4) — raw relative light units (RLU) from the
plate reader, plus relative values (percent of the 24 h mean or of control) on the
Replicates_C2(Fig3C) andCycle_3(FigS4) sheets; the instrument metadata (Tecan Spark) is retained on the
Result sheet. - Immunoblot densitometry (FigS11) — background-subtracted band intensities for
Anti-Halo(endogenously tagged GlRac) andTubulin(loading control), the
Halo/TubulinRatio, and the ratio normalized to the paired control, for each of two
knockdown/control pairs. - Colocalization / signal-coincidence scoring (Fig5) — per stage (
Cup/Hollow/
Full), the number of cells and whether the marker signal coincided with the
GlRac compartment (Colocalize?).
Conventions: timepoints are in hours; dates embedded in filenames are YYYYMMDD (or MMDD
with the year given by the folder/context); the decimal separator is a period; empty cells
denote no measurement/not applicable. Replicates correspond to independent biological
experiments performed on different dates.
Folder contents (per figure)
Fig1 — encystation induction. Fig1B_* three replicate workbooks scoring compartment
number over 0–4 h encystation, plus the Prism summary; Fig1D_* compartment sizing (µm).
The 0 h encystation data share measurements with the 48 h time point in Fig2: the 0 h
sheets of the 20240313 and 20240320 Fig1B workbooks are the same counts as the 48 h sheets
of the 20240313 and 20240320 Fig2B workbooks, and all 126 Fig1D 0 h diameters are the
48 h diameters in Fig2D_Compartment_sizing_48_72_96h (sheet 48hr PE).
Fig2 — nutrient availability. Fig2B_* compartment number during normal growth (PE,
3 reps + Prism); Fig2C_* compartments per cell at 48/72/96 h; Fig2D_* sizing at
48/72/96 h; Fig2F_* counts after refeeding with fresh TYDK (0–3 h, 3 reps); Fig2G_*
individual-nutrient refeeding (glucose / casein digest / L-arginine, 3 reps).
Fig3 — GTOR. Fig3C_* GTOR-NLuc luminescence at 24 h vs 72 h in PE (raw RLU +
relative); Fig3D_* proportion of positive cells after GTOR morpholino knockdown (four
experiments in three workbooks; rep2_rep3 holds two);
Fig3E_* compartments per cell (GTOR KD, 4 reps); Fig3F_* compartment sizing (GTOR KD).
Fig4 — time-lapse videos. Three subfolders: ...forming_Dorsal_area and
...forming_Ventral_area (compartment biogenesis) and ...resolving (compartment
clearance: a 4 h 2 min example, two labelled clips of individual cells
(cell2_slice9_... and cell3_slice8_...), and nine montage GIFs). These are the source
movies for the S1–S3 Movies described in the article.
Fig5 — conserved features. Fig5D/E/F_* concanamycin A (0.25 µM) effect on proportion,
compartments per cell, and size; Fig5H_CatL_9548-mNG.pdf cathepsin-L localization;
Fig5I_* acidification (LysoTracker Deep Red) and cathepsin/actin signal-coincidence
scoring, with Prism summary.
Fig6 — cysteine proteases / E-64d. Fig6A-B_Extra_examples.pdf representative
BODIPY-LHVS images; Fig6C_* proportion of positive cells (E-64d vs vehicle, 3 reps);
Fig6D_* compartments per cell; Fig6E_* compartment sizing.
Fig7 — GlRac knockdown and constitutive activation. Fig7B_* proportion of positive
cells after GlRac KD (DALGreen, 3 reps); Fig7C/D_* compartments per cell and size (KD);
Fig7F_* CA-Rac (tet-induced, cathepsin-labeled, 3 reps); Fig7G/H_* compartments per cell
and size (CA-Rac).
Fig8 — quinacrine. Fig8D_* proportion of positive cells (10/15 µM quinacrine vs
vehicle, 3 reps); Fig8E_* compartments per cell; Fig8F_* compartment sizing.
FigS3 — rapamycin. Rapamycin (36 µM, 2 h) vs DMSO: per-image scoring workbooks for two
experiments (20241121, 20241127) + Prism summary of three experiments (see above).
FigS4 — GTOR morpholino validation. GTOR-NLuc knockdown NanoLuc data + Prism (GTOR-NLuc reduced to
~30.1% of control, a ~69.9% reduction).
FigS6 / FigS7 — ATG candidate localization. Representative micrographs (.pdf) of
mNG-tagged candidates, each named by its GiardiaDB number. S6: GlATG1.1-like (137719),
GlATG1.2-like (17566), GlVPS15-like (113456), GlVPS34-like (17406), GlATG16-like
(33762), GlATG21-like (16957), and GlATG18-like (10822). S7: proposed ATG8 (28994) and
ATG7 (6288) candidates.
FigS10 — additional cathepsins. Micrographs of cathepsin B (29304) and cathepsin
L-like (137680, 3099) mNG fusions.
FigS11 — GlRac immunoblot. Quantification.csv / Quantification.xlsx densitometry
(anti-Halo, tubulin, ratios; 70.3% reduction; see "Cells outside the main data table"); WB_membrane.mscn the original Bio-Rad Image Lab scan; and
WB_membrane_ch0_Red.tif / WB_membrane_ch1_Green.tif, the two acquisition channels
exported as open 16-bit grayscale TIFFs (2750 × 2200 px) for viewing without Bio-Rad
software.
FigS12 — knockdown penetrance. Per-replicate DALGreen scoring of GlRac KD penetrance
(3 reps) + Prism. These files carry the prefix FigS12B.
Code/Software
No custom analysis code was used.
- Image acquisition/deconvolution: DeltaVision Elite deconvolution microscope (GE),
60×/1.42 NA objective, PCO Edge 5.4 sCMOS camera; SoftWoRx (Applied Precision). - Image measurement: ImageJ/Fiji v1.54f (compartment sizing and processing).
- Luminescence: Tecan Spark plate reader (NanoLuc GTOR assays).
- Statistics and graphing: GraphPad Prism 10 for Windows (v10.5.0). Paired two-tailed
t-tests for matched-sample comparisons; two-tailed Welch's t-tests for compartments-per-
cell and size comparisons (violin plots). P ≤ 0.05 considered significant. - Figure assembly: Adobe Illustrator CS6.
The .prism files require GraphPad Prism to open; all underlying values are duplicated in
the accompanying .csv files (and the original Excel workbooks), so the data remain usable
without proprietary software. The exceptions are the third FigS3 rapamycin experiment and
the Fig5I DALGreen values, which are recorded only in the Prism files.
Access information
- Other publicly accessible location of these data: the associated journal article and its
Supporting Information (add DOI on acceptance). - Data derived from external resources: gene identifiers use GiardiaDB
GL50803_*
accessions; the G. lamblia WB clone C6 strain is ATCC 50803.
License
All files in this deposit are released under the Creative Commons Zero (CC0 1.0)
public-domain waiver, per Dryad requirements. No included file carries a conflicting license.
