Cell lines and cell culture conditions

The HBEC3-KT cell line (RRID:CVCL_X491) was a gift from Dr. John Minna (The University of Texas Southwestern Medical Center, Dallas, TX) and was cultured in Keratinocyte serum-free medium (K-SFM; Gibco, 17,005,042) with supplied recombinant epidermal growth factor and bovine pituitary extract added. The KRAS and shLKB1 HBEC3-KT cell lines were produced as described previously (Koo et al. 2024). The LKB1-wildtype and LKB1-knockout HBEC3-KT clonal populations were produced by Synthego, in consultation with the Emory Integrated Genomics Core facility, using their antibiotic-free CRISPR/SpCas9 single guide knockout method. The guide RNA sequence used was CGAUGAGCUUGGCCCGCUUG.

The A549 cell lines (RRID:CVCL_0023) were a gift from Dr. Wei Zhou (Winship Cancer Institute of Emory University, Atlanta, GA) and produced as previously described (Liu et al. 2015). The H1299 cell lines (RRID:CVCL_0060) were a gift of Dr. Adam Marcus (Winship Cancer Institute of Emory University, Atlanta, GA) and produced as previously described (Konen et al. 2017). The JK-43-M cell line was produced in our lab as previously described (Koo et al. 2024). The A549, H1299, and JK-43-M cell lines were maintained in RPMI 1640 medium with L-glutamine (Corning, 10–040-CV) supplemented with 10% fetal bovine serum (FBS; Corning, 35–011-CV) and 1% PenStrep-glutamine (Gibco, 10,378,016). For experiments, cells were cultured in RPMI-1640 with 10% FBS and without antibiotics.

All cell lines were tested and confirmed negative for mycoplasma (Universal Mycoplasma Detection Kit, ATCC, 30-1012 K) and used for experiments within three months of thawing.

Glucose starvation assay of LKB1 activity

Though other kinases, such as CAMKK2, can phosphorylate AMPK at the T172 position, only LKB1 phosphorylates AMPK at T172 in response to glucose starvation; therefore, a glucose starvation assay can be used to test whether LKB1 is functional in a cell line. We cultured cells in their typical complete culture media and plated them onto 10-cm dishes and allowed them to grow until approximately 90% confluence. Complete medium was aspirated and replaced with either RPMI with 2 mg/L glucose (Corning, 10–040-CV) or RPMI without glucose (Corning, 10–043-CV); of note, these media were not supplemented with FBS. Samples were collected after 4 h of incubation in a cell culture incubator (21% O2, 5% CO2, 37 °C) and processed for western blot as described below.

siRNA knockdown of PER1

Dharmacon ON-TARGETplus siRNA SMARTPool reagents were used. The siPER1 SMARTPool (Horizon Discovery, L-011350–00-0005) targeted the following sequences: CCAAUAAGGCGGAGAGUGU, CCAGUGACCUGCUCGAACU, GGCCGAAUCGUCUACAUUU, and CAACGGGCAUGAGUCUAGA. As a negative control, the ON-TARGETplus non-targeting control pool siRNA (siNTC) was used (Horizon Discovery, D-001810–10-05) and targeted the following sequences: UGGUUUACAUGUCGACUAA, UGGUUUACAUGUUGUGUGA, UGGUUUACAUGUUUUCUGA, and UGGUUUACAUGUUUUCCUA. The transfection reagent used was DharmaFECT 1 (Horizon Discovery, T-2001–02). Transfections for all cell lines were conducted using the reagent concentrations provided in the Dharmacon protocol for A549 cells.

Invasion assays

HBEC3-KT 3D invasion assays were performed as previously described (Koo et al. 2024). H1299 3D invasion assays and image analysis were performed as previously described (Konen et al. 2017), with cells treated with siRNA 24 h prior to spheroid formation via centrifugation of cells in ultra-low attachment 96-well plates (Corning, 7007).

Western blot analysis and antibodies

To prepare whole cell lysates, cells were washed with ice-cold phosphate-buffered saline (PBS), then scraped into 1 ml of fresh ice-cold PBS and transferred into a 1.5 ml tube. Cells were pelleted by centrifugation (2000 × g for 5 min at 4 °C) and the cell pellet was then resuspended in high-salt NP-40 lysis buffer (Alfa Aesar, J61428) supplemented with a protease and phosphatase inhibitor cocktail (Cell Signaling, 5872), incubated on ice for 10 min, then briefly sonicated to shear DNA. To prepare subcellular fractionation lysates, the REAP method (Suzuki et al. 2010) was used with 100 μl volume in each fraction, sonication of all samples, and addition of protease and phosphatase inhibitor cocktail (Cell Signaling, 5872) to finished fractions. Protein concentration was measured using Bradford assay (Bio-Rad, 5,000,205) and samples with equal protein concentrations were prepared using homemade 5X Laemmli buffer supplemented with 10% beta-mercaptoethanol and then boiled for 5 min. After electrophoresis, gels were transferred overnight onto PVDF membrane (Bio-Rad, 1,620,177). Membranes were blocked in TBS-T containing 5% non-fat milk powder (Cell Signaling, 9999; Apex, 20–241) and 5% BSA (Cell Signaling, 9998).

The following primary antibodies were diluted in 5% BSA in tris-buffered saline (Cell Signaling, 12498S) supplemented with 0.01% Tween-20 (TBS-T; ChemCruz, sc-29113B) and used at the indicated dilutions: PER1 (1:1000, Boster Bio, A00876, RRID:AB_3086701), LKB1 (1:1000, Cell Signaling, #3050, RRID:AB_823559), JAG1 (1:1000, Cell Signaling, #70,109, RRID:AB_2799774), pAMPK-T172 (1:1000, Cell Signaling, #2535, RRID:AB_331250), AMPK (1:1000, Cell Signaling, #2532, RRID:AB_330331), Actin (1:500, DSHB, JLA20, RRID:AB_528068; 1:1000, Sigma-Aldrich, A5441, RRID:AB_476744), Lamin A/C (1:1000, DSHB, MANLAC3(4C10), RRID:AB_2618205). The following HRP-conjugated secondary antibodies were diluted in 5% non-fat milk in TBS-T at a 1:3000 dilution: anti-Rabbit (Boster Bio, BA1054, RRID:AB_2734136) and anti-Mouse (Fisher Scientific, 62–652-0, RRID:AB_2533947). Blots were stripped after each successive primary antibody using OneMinute Plus Western Blot Stripping Buffer (GM Biosciences, GM6011). Blots were developed using SuperSignal West Pico PLUS chemiluminescent substrate (Thermo Scientific, 34,580) and imaged using a BioRad ChemiDoc.

Cell proliferation assays

Cell proliferation assays were completed using two complementary methods: Cell Counting Kit-8 (CCK-8; Dojindo, CK04) and by manually counting trypsinized cells using a hemocytometer. For the CCK-8 method, after siRNA treatment HBEC3-KT cells were seeded in quadruplicate into multiple 96-well plates, one plate for each timepoint, at a density of 2000 cells per well and a volume of 100 μl per well. At each timepoint, 10 μl of CCK-8 reagent was added to each test well, incubated for 2 h, and then 450 nm absorbance was read using a microplate spectrophotometer (Biotek Epoch). For each condition, a control well without CCK-8 added was used to subtract background absorbance, leaving 3 wells as technical replicates.

For the cell counting method, cells were seeded, with HBEC3-KTs at a density of 2.5 × 104 cells per well and A549s at a density of 2 × 104 cells per well, in 24-well and 12-well plates, respectively. For the JK-43-M cells, 1 × 105 cells were seeded into 6-well plates. At each timepoint, cells were trypsinized and resuspended in RPMI supplemented with 10% FBS to inactivate the trypsin, taking note of the total volume. Cells were counted using a hemocytometer to calculate total cell number per well.

RNA-seq and gene expression analysis

RNA-seq experiments and analysis of H1299 parental, leader, and follower cells (Zoeller et al. 2019) and HBEC3-KT spheroids (Koo et al. 2024) were performed as previously described. For RNA-seq of the LKB1-WT and LKB1-KO HBEC3-KT cells, cells were plated in triplicate on 150 mm cell culture dishes and synced with a 30-min pulse of 100 nM dexamethasone. Samples were collected at 4 h increments from 24 to 48 h post-syncing. At each timepoint, the adhered cells were washed with ice-cold PBS, collected by scraping into ice cold PBS, pelleted, then flash frozen and kept at -80 °C until processing by the Emory Integrated Genomics Core. Total RNA was isolated using the miRNeasy Mini Kit (Qiagen, 217,004) then quantified and quality-controlled using the 2100 BioAnalyzer (Agilent). RNA-seq was performed by DLS Hudson Alpha at a sequencing depth of 50 million paired-end reads. For data processing, sequence adaptors were trimmed from fastq files using cutadapt (RRID:SCR_011841) and TrimGalore-0.6.10 (RRID:SCR_011847). Sequences were then aligned to the Homo sapiens reference genome release 111 using STAR genome aligner for paired end reads (RRID:SCR_004463). The resulting gene count data were normalized and differential expression analysis performed using DESeq2 (RRID:SCR_015687). QC inspection of the data resulted in one sample (KO_24_3) being removed from analysis. All timepoints were pooled for the analysis presented here.

MCP-counter analysis of intratumoral immune cells

Full datasets of normalized gene expression for the TCGA-LUAD and CPTAC3 studies were downloaded from the Genomic Data Commons Data Portal (RRID:SCR_014514) for further analysis. The MCPcounter R package (Becht et al. 2016) was installed from GitHub using the devtools package and used with default gene signatures. After calculation of MCP-counter scores using R version 4.4.1, the data from each study were subset to only the samples present in the PER1/PER1 expression quartiles, sorted by group, and the data were then statistically analyzed by two-way ANOVA followed by Šídák's multiple comparisons test in GraphPad Prism 10.4.1.

Statistical analysis

Aside from statistical test results for clinical data calculated within cBioPortal, or others as indicated, all other statistical analyses were performed using GraphPad Prism 10.4.1 (627) (RRID:SCR_002798).

Sex as a biological variable

Both male and female specimens from cBioPortal datasets were included in the analysis, but sexes were not analyzed separately or compared due to limited sample size and statistical power.

Data availability statement

The RNA sequencing data analyzed in this study are available as follows: 3-D spheroid HBECs (NCBI Gene Expression Omnibus, GSE271368); H1299 parental, leader, and follower subpopulations (NCBI Sequence Read Archive, PRJNA542374); and synced time course of LKB1-WT and LKB1-KO cells (NCBI Sequence Read Archive, PRJNA1258988). Gene expression, protein abundance, mutation, and clinical data from human lung cancer specimens from the TCGA PanCancer Atlas, CPTAC 2020, and CPTAC GDC (July 2024) studies were accessed and downloaded through cBioPortal (RRID:SCR_014555, https://www.cbioportal.org/). Other data files will be provided upon request.