**4. Materials and Methods**

#### *4.1. Cell Culture*

Cells were cultured in DMEM media (Gibco Cat # 11885, Billings, MT, USA) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin.

### *4.2. Cell Cycle Analysis*

Cell cycle analysis was performed using propidium iodide staining of fixed MDA-MB-231 cells treated with ZJ-101. Briefly, a 70% confluent 10 cm dish of MDA-MB-231 was collected via trypsinization and washed with 1× PBS. Pelleted cells were fixed by the dropwise addition of 2 mL of ice-cold 75% EtOH. Fixed cells were washed again with PBS and stained with 1 mg/mL propidium iodide solution prior to FACS analysis.

#### *4.3. Growth-Rate Inhibition Metric Analysis*

Control populations of MDA-MB-231 cells were prepared for growth rate normalization at seeding densities of 50, 200, 500, and 1000 cells per well in a 384-well flat bottom plate (Corning Inc., Corning, NY, USA). Following compound treatment, cells were stained with 1 μM Calcein AM (for live cells), 20 μg/mL Hoechst 33342 (for nucleus), and 1 μM ethidium homodimer (a cell membrane impermeable dye for dead cells) for 15 min at 37 ◦C prior to imaging. Imaging was performed using the 4× objective on the ImageXpress Micro (Molecular Devices, San Jose, CA, USA) with image-based focusing. Collected images were assessed using the live/dead program in MetaXpress (version 6.1), and data were organized and entered into grcalculator.org, where growth rate normalized inhibition calculations were performed for each compound tested.

#### *4.4. 3D Spheroid Assays*

Three-dimensional spheroid formation was assessed using Corning ultra-low attachment (ULA) plates. MDA-MB-231 cells were seeded in ULA plates at a density of <sup>2</sup> × 103 cells/well (96-well) or 5 × 102 cells/well (384-well) in DMEM media supplemented with 10% FBS and centrifuged for 5 min at 400× *g*. Cells were then left undisturbed in a 37 ◦C, 5% CO2 incubator for 72 h to form spheroids. ZJ-101 was added at the indicated concentrations and incubated for a further 72 h. Pretreatment of ZJ-101 was also performed with basic light images taken every 24 h for up to 72 h. Spheroidicity was determined based on the overall diameter normalized to an untreated control using ImageJ (version 1.52r).

The 3D spheroid combination assay was performed as a spheroid disassembly assay requiring 72 h of pre-formed spheroids prior to a further 72 h of drug addition. ZJ-101 and etoposide were arranged with 1:3 dilutions in an 8 × 8 matrix format at the indicated concentrations. Fluorescent images were acquired after staining using the same protocol and instrumentation outlined in the GR assay. Laser-based focusing was utilized to obtain clear spheroid images from the bottom of each well. Ten z-stack images were taken and combined into a single maximum-intensity image. Images were processed with the live/dead program in MetaExpress.

#### *4.5. Cell Staining and Imaging*

Basic light microscopy was performed on HEK293T cells cultured in 6-well TC-treated plates (Greiner #657160, Kremsmünster, Austria) by capturing images through the 20X objective of a Zeiss Axiovert 25 (Carl Zeiss Microscopy, White Plains NY, USA) with a 12.2 megapixel camera. For MDA-MB-231 spheroid light microscopy, the same protocol was used with a 10X objective.

For all staining assays, HeLa cells were cultured on 96-well CellView TC-treated microplates (Greiner #655891, Kremsmünster, Austria). Cells were washed with cold PBS prior to fixation with 4% PFA for 10 min at RT. After fixation, cells were washed twice with PBS and permeabilized with 0.1% Triton-X for 10 min at RT. Cells were again washed twice with PBS and blocked with cell staining buffer for 30 min. Antibodies for GM130 (Cell Signaling #12480, Danvers, MA, USA) were added at 1:1000 in staining buffer for 1 h at RT. After three washes with PBS, secondary anti-rabbit AlexaFluor-488, or AlexaFluor-647 (ThermoFisher, Waltham, MA, USA) were added along with the indicated fluorescent lectin conjugates (HPA-647, WGA-488, and PNA-555 from ThermoFisher) at 1:1000 dilution for 1 h RT. Cells were again washed and stained with 1:10<sup>4</sup> Hoechst 33,342 for 3 min prior to imaging with an OperaPhenix (PerkinElmer, Waltham, MA, USA). Images were uploaded to the Columbus Analyzer (version 2.9.1.699) and processed for high-content analysis.

For the live cell lysotracker assay, HeLa cells were cultured as above during treatment with the specified compounds. Cells were loaded with DMEM media containing 100 nM Lysotracker Deep Red for 30 min at 37 ◦C. After lysotracker staining, media was exchanged for Live Cell Imaging Solution (Invitrogen #A14291DJ, Waltham, MA, USA) containing 1:10<sup>4</sup> Hoechst 33342 for nuclei staining. Images were uploaded to the Columbus Analyzer and processed for high-content analysis using default settings for spot detection and intensity calculation.

#### *4.6. Transcriptome Analysis*

MDA-MB-231 cells from 10 cm dishes were harvested through scraping (for 2D) or spheroids collected (for 3D) using a wide-gauge pipette and subjected to RNA extraction via the RNeasy mini kit using the manufacturer's instructions. Biological replicates of N = 3 were used for both sets of analyses, with N = 96 spheroids representing a single biological replicate for the 3D populations. RNA-sequencing was performed by Genewiz as paired-end 150 bp reads following poly-A selection to enrich mRNA transcripts. Paired-end FASTQ files were uploaded to Galaxy using the public server at usegalaxy.org (accessed on 7 April 2021) and aligned to hg38 using HISAT2. Transcripts were assembled and counted using htseq-count, and differential gene expression was evaluated with DESeq2 with default settings. Fold changes were assessed against the DMSO vehicle controls. Transcripts were annotated using the most current GENCODE release. Gene Ontology analysis was performed using g:Profiler at https://biit.cs.ut.ee/gprofiler/gost (accessed on 8 April 2021) by inputting the top 200 significant genes for each concentration of ZJ-101 and arranging them by descending order of significance [9]. Heatmaps and PCA plots were generated using ClustVis v1.0 at https://biit.cs.ut.ee/clustvis/ (accessed on 8 April 2021) [10].

The data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus and are accessible through GEO accession number GSE231359 https://www.ncbi. nlm.nih.gov/geo/query/acc.cgi?acc=GSE231359 (accessed on 8 April 2021) [31].

#### *4.7. Glycan Analysis*

#### 4.7.1. Glycan Incorporation Assay

Briefly, 50 μM azido-modified sugars (Invitrogen) tetraacetylated N-azidoacetylglucosamine (GlcNAz), tetraacetylated N-azidoacetylgalactosamine (GalNAz), tetraacetylated N-azidoacetyl-D-mannosamine (ManNAz), and 100 μM alkynyl-fucose (Invitrogen) were added to 96-well plates in combination with the indicated doses of ZJ-101. Cells were then washed, fixed, and permeabilized prior to the click reaction. Copper-catalyzed click reactions were performed using the Click-iT Cell Reaction Buffer Kit (Invitrogen C10269) per the manufacturer's instructions, containing 5 μM of either Fluor alkyne-647 (Invitrogen A10278) to label azido-incorporated sugars or Fluor Azide-488 to label incorporated alkynylfucose. Plates were washed five times before Hoechst counterstaining and imaging. Glycan incorporation was determined by the total intensity of each signal normalized to untreated control cells.

#### 4.7.2. Glycome Profiling

Glycomics profiling was performed by Creative Proteomics. N-glycans were prepared from fresh cell pellets washed with PBS, resuspended in 1 mL of lysis buffer, and sonicated (5 pulses of 10 s). Samples were then dialyzed in 50 mM ammonium bicarbonate for 24 h at 4 ◦C, with the buffer changed three times. Following dialysis, the samples were lyophilized. To the lyophilized powder, 1 mL of 2 mg/mL DTT was added, and the solution was incubated at 50 ◦C for 2 h. Briefly, 1 mL of a 12 mg/mL IAA (iodoacetamide, Sigma, St. Louis, MO, USA) solution was then added and incubated at RT in the dark for 2 h. The DTT and IAA-treated proteins were then dialyzed against 50 mM ammonium bicarbonate (Sigma) for 24 h at 4 ◦C. Samples were next resuspended in 1 mL of 500 μg/mL TPCKtreated trypsin (Sigma) solution and incubated at 37 ◦C overnight. The trypsin reaction mixture was purified over C18 Sep-Pak columns (Waters, Milford, MA, USA) by 1-propanol elution. Fractions containing peptides were pooled and lyophilized. The lyophilized peptides were resuspended in 200 μL of 50 mM ammonium bicarbonate, to which 3 μL of PNGaseF (New England Biolabs, Ipswich, MA, USA) was added for a 4 h incubation at 37 ◦C. Following this initial incubation, another 5 μL of PNGaseF was added for overnight incubation at 37 ◦C. The enzymatic reaction was stopped by the addition of two drops of 5% acetic acid, and the released N-glycans were purified over C18 Sep-Pak columns. Flow-through and 5% acetic acid washing fractions containing the released N-glycans were pooled and lyophilized and were subject to permethylation. For O-glycan analysis, the PNGaseF-treated glycopeptides were eluted from the C18 column with 1-propanol. The lyophilized eluted peptides were subjected to *O*-glycan preparation.

*O*-glycan-containing powder was solubilized by 400 μL of a sodium borohydride (Sigma-Aldrich) solution in 0.1 M NaOH (55 mg NaBH4/1 mL 0.1 M NaOH) and incubated overnight at 45 ◦C. The reaction was stopped by adding drops of pure acetic acid until the fizzing stopped. The samples were passed through a Dowex 50W X8 resin (Sigma-Aldrich) column, and the pooled fractions were dried by lyophilization. The lyophilized samples were next resuspended in 1 mL of an acetic acid:methanol solution (1:9 *v*/*v*) and co-evaporated under nitrogen flow. This step was repeated 3 more times, and the dried samples were resuspended in 200 μL of 50% methanol prior to being loaded onto the C18 Sep-Pak column. Free *O*-glycans were collected in the flow-through and 5% acetic acid wash fractions. These fractions were pooled, lyophilized, and subjected to permethylation.

Permethylation was performed as follows. Seven pellets of NaOH in 3 mL of DMSO were ground with a mortar and pestle. One milliliter of this slurry solution was added to the dry sample in a glass tube with a screw cap. Five hundred microliters of iodomethane was then added to the slurry and shaken at RT for ~30 min. After the reaction reaches completion, noting the formation of white solids, the cap is released slowly to relieve the gas pressure that has built up. One milliliter of MilliQ water was added to stop the reaction, and the sample was vortexed until all solids were dissolved. To the sample, 1 mL of chloroform and an additional 3 mL of MilliQ water were added with continuous vortexing to mix both phases. The samples were then centrifuged briefly to separate the chloroform and the water phases (~5000 rpm, <20 s). The aqueous top layer was discarded, and washing was repeated twice with the addition of 3 mL of MilliQ water. The chloroform fraction was then dried with a SpeedVac (~20–30 min). A C18 Spe-Pak (200 mg) column was prepared with methanol, MiliQ water, acetonitrile, and MilliQ water. The dry sample was resuspended with 200 μL of 50% methanol and loaded onto the column. The column is then washed with 2 mL of 15% acetonitrile and eluted into a clean glass tube with 3 mL of 50% acetonitrile. Finally, the eluted fraction was subjected to MS analysis.

MS data were acquired on a Bruker UltraFlex II MALDI-TOF mass spectrometer instrument (Bruker Scientific LLC, Billerica, MA, USA). Reflective positive mode was used, and data were usually recorded between 500 *m*/*z* and 6000 *m*/*z* for *N*-linked glycans and between 500 *m*/*z* and 4000 *m*/*z* for *O*-glycans. For each MS N- and *O*-glycan profile, the aggregation of 20,000 laser shots or more was considered for data extraction. Mass signals with a signal/noise ratio of at least 2 were considered, and only MS signals matching

an *N*- and *O*-glycan composition were considered for further analysis and annotation. Subsequent MS post-data acquisition analysis was made using mMass [32].
