Breast cancer progression involves extensive remodeling of the extracellular matrix (ECM), including increased stiffness, altered viscoelasticity (stress relaxation), and elevated collagen levels. While in vitro experiments have revealed a role for each of these factors in individually promoting malignant behavior, their combined effects remain unclear. Here, we engineered alginate-collagen hydrogels with independently tunable stiffness, stress relaxation, and collagen density to dissect how the complex ECM environment regulates cancer cell phenotype. We show that high stiffness, fast stress relaxation, and high collagen density led to changes in cell morphology, marked by decreased roundness, and promoted spheroid invasion in both breast cancer and non-transformed mammary epithelial cells. Single cell migration speed and displacement were greatest in matrices of high stiffness, low collagen density, and slow stress relaxation. RNA-seq and Cleavage Under Targets and Tagmentation (CUT&Tag)-seq revealed that high stiffness and fast stress relaxing groups were enriched for Sp1 target gene expression as well as increased Sp1 binding at genomic loci. Notably, analysis of publicly available claudin-low breast cancer data showed that high expression of the Sp1-regulated genes in fast stress relaxing groups was correlated with significantly reduced patient survival. Mechanistically, we found that phosphorylated Sp1 (T453) exhibited increased nuclear localization in matrices with high stiffness and fast stress relaxation. Furthermore, Sp1 phosphorylation was regulated by PI3K and ERK1/2 activity, as well as actomyosin contractility. Our tunable hydrogel platform reveals that multiple tumor-mimicking cues within complex viscoelastic microenvironments reinforce malignant traits, with Sp1 acting as a mechanoresponsive transcription factor that transduces these signals.

RNA-seq dataset for MDA-MB-231 and MCF-10A cells encapsulated for 7 days in soft or stiff, slow or fast relaxing, and low or high collagen density alginate-collagen matrices. The RNA-seq dataset has 8 total mechanical conditions containing 3 replicates per condition for MDA-MB-231 cells and 2 replicates per condition for MCF-10A cells. The dataset contains both fastq and raw counts (.xlsx) files. 

The CUT&Tag-seq dataset was generated for Sp1 in MDA-MB-231 cells encapsulated in soft or stiff, slow or fast relaxing, low collagen density alginate-collagen matrices. The dataset contains has 4 total mechanical conditions and 2 replicates per condition. This dataset contains the fastq and bigwig (.bw) files. 

For RNA-seq, cells were extracted from alginate-collagen matrices by rocking them in Falcon tubes containing 2.5 mg/ml collagenase (Sigma, Cat#C0130) solution in PBS for 30 minutes at 37°C. After this, the tubes were centrifuged, and the supernatant was removed. Following this, the pellet was dissolved in 10 ml ice-cold EDTA (50 mM) and placed on a rotator for 10 mins. The tubes were centrifuged, and the supernatant was then removed. The cell pellet was then lysed using Trizol (Life Technologies), and RNA was extracted using the total RNA mini prep kit according to the manufacturer’s instructions (Epoch Life Sciences).

Bulk-mRNA-sequencing library prep was done using the Cel-Seq2 pipeline as described previously. Briefly, 10 ng RNA was reverse-transcribed using the CelSeq2 RT-primer, DTT (0.1 M), dNTPs (New England Biolabs, Cat# N0447l), and Superscript II reverse transcriptase (Invitrogen, Cat# 18064014), followed by second-strand synthesis using RNAseH (ThermoFisher Scientific, Cat# EN0202), E.Coli DNA Pol I (Invitrogen, Cat# 18010025), E.Coli ligase (Invitrogen, Cat#18052-019), and the second strand buffer (Invitrogen, #18052-019). After cDNA cleanup using DNA beads (AMPure, Cat# A63882), in vitro transcription was performed using the MEGAscript T7 kit (Invitrogen, Cat# A57622), resulting in amplified single stranded RNA (aRNA). After aRNA treatment with ExoSAP-IT PCR reagent (ThermoFisher, 78200), RNA strands were fragmented using a fragmentation buffer containing 200 mM Tris-acetate (pH 8.1), 500 mM KOAc, and 150 mM MgOAc. aRNA cleanup was then performed using RNA-beads (AMPure, Cat#A63987). Next, aRNA was reverse transcribed using a RT random hex primer, and cDNA was then amplified using an RNA PCR primer (RPI), a uniquely indexed Illumina primer, and a PCR master mix (New England, Cat#M0541S). After bead cleanup and quality control using a bioanalyzer (Agilent), DNA libraries were sequenced using the NovaSeq™ 6000 and X platforms (Illumina). The raw reads were then mapped to the hg19 (GRCh37) reference genome.

For CUT&Tag-seq, 250,000 cells per sample were washed twice using 1 ml wash buffer (also containing protease inhibitor) at room temperature. A Concanavalin A bead slurry was prepared by dissolving the beads in 1X binding buffer and separating on a magnetic stand. After repeating this step once, the beads were resuspended in 20 μl of binding buffer and then slowly added to the cell suspension and rotated for 10 minutes. Next, after removing the supernatant on a magnetic stand, cells were suspended in 50 μl of ice-cold Sp1 antibody buffer containing 5% digitonin, protease inhibitor, and Sp1 primary antibody (Sigma, cat # 07645, 1:50 dilution) and left on a rotator overnight at 4 °C. The next day, after removing the supernatant on a magnetic stand,100 μl of diluted anti-rabbit secondary antibody (1:100) in Dig-Wash buffer (also containing protease inhibitor and 5% digitonin) was added to each sample. Tubes were placed on a rotator for 60 minutes at room temperature. After removing the supernatant, cells were washed 2X using the Dig-Wash buffer. Next, the assembled pA-Tn5 transposome enzyme was diluted in the Dig-300 buffer (also containing protease inhibitor) at a 1:100 concentration and incubated for 60 minutes at room temperature. After washing with 1 ml Dig-300 buffer twice, 125 μl tagmentation buffer (also containing protease inhibitor and digitonin) was added to each sample and incubated for 60 mins at 37°C. Tagmentation was then immediately stopped by adding 4.2 μl of 0.5 M EDTA, 1.25 μl of 10% SDS, and 1.1 μl of Proteinase K (10 mg/ml) to each sample and incubated at 55°C for 60 minutes. After this, the supernatant was collected using a magnetic stand, and 625 μl DNA purification binding buffer was added to each tube, and the mixture was transferred to a DNA column. After spinning down tubes at 17,000 g for 1 minute, the flow through was discarded, and the column was washed at 17,000 g, using DNA purification wash buffer (also containing 80% ethanol). The DNA was eluted in 35 μl of nuclease-free water. Next, PCR amplification was performed using the Illumina i7 and i5 indexed primers, dNTPs (New England Biolabs, Cat# N0447l), and PCR master mix (New England, Cat# M0541S). The size distribution of the libraries was quantified on an Agilent Bioanalyzer 2100 and Qubit fluorometer. Libraries were subjected to paired-end 150 bp Illumina sequencing on a Novaseq X platform.