335 patients were provided by the First Affiliated Hospital of Chengde Medical University who suffer from malignancy. There are 332 (99.1%) valid questionnaire and 3 (0.9%) invalid questionnaire. Inclusion criteria included: (1) Patient were pathologically diagnosed with cancer; (2) 18-80 years of age; (3) Patient had not taken antidepressants; (4) Patient had not any condition that would impede understanding of the study or the informed consent. Exclusion criteria included: (1) Patients with severe mental illness and cognitive dysfunction; (2) Patients who refuse to participate in the research and cannot communicate with the researchers normally; (3) The expected life is less than 3 months; (4) Patients with hearing and language impairment. The results of Self-Rating Anxiety and Self-Rating Depression in 215 patients with different types of cancer are available in Table S1 . 215 blood samples were collected with the consent of the patients from the above 332 patients.

The Zung Self-Rating Anxiety Scale (ZSAS) was adopted for assessing the anxiety severity, whereas the Zung Self-Rating Depression Scale (ZSDS) was employed for evaluating depression severity. The SAS and SDS index scores are then categorized into 2 levels respectively: no significant psychopathology (SAS Index <50 and SDS Index <53); presence of varying degrees of anxiety (SAS Index ≥50) or depression (SDS Index ≥53). The Zung score does not necessarily give the clinical diagnosis of anxiety or depression, instead, it suggests the severity of the clinically significant symptoms.

Human umbilical vein endothelial cells (HUVECs), as well as human gastric cancer cell lines MGC803, were provided by Chinese Academy of Military Medical Sciences. All these cells were cultivated within the RPMI-1640 medium that contained 10% FBS. hCMEC/D3 was purchased from SCIENCE CELL Research Laboratories. hCMEC/D3 cells were cultivated within the Endothelial Cell Medium (ECM, Cat.No.1001). HUVECs and hCMEC/D3 were treated with β2-AR agonist, Isoprenaline (ISO) and β2-AR blocker, ICI 118,551 hydrochloride (ICI, Sigma-Aldrich, St. Louis, MO, USA).

Liposomes shRNA vectors carrying the shRNA sequences were purchased from Gene Pharma (Shanghai, China), the sequence is in Table 1 . HUVECs were transfected with shplexinA1 and shNC (Negative control) with Polyplus Transfection (Strasbourg, France).

HUVECs were seeded at a density of 1 × 10⁴ cells/well in a 96-well flat-bottom plate and treated with different concentrations of ISO and ICI. HUVEC proliferation was measured using CCK8 cell proliferation assay kit (Beijing Solarbio Science & Technology Co., Ltd., china) according to the manufacturer’s specifications after 0, 6, 12, 24 and 48 h of incubation, and the absorbance was measured at 450 nm using a microplate reader.

The RNeasy Mini Kit (Invitrogen) was utilized for extracting total RNA. Then, 2 µg total RNA extracted was collected to prepare cDNA by the use of M-MLV reverse transcription system kit (Promega, Madison, WI, USA). All PCR primers sequences are available in Table 1 . The SYBR Green qPCR Detection System (Promega, Madison, WI, USA) was utilized for qRT-PCR following specific protocols. The comparative threshold cycle approach was employed for result analysis, and GAPDH was used to be the internal reference. In the meantime, the 2^-ΔΔCt formula was adopted to determine fold change (FC) in mRNA expression.

The cold RIPA buffer (Thermo Scientific, Rockford, IL, USA) was utilized for cell lysis. Minute (TM) Cytoplasmic and Nuclear Fractionation kit (SC003) was used to extract nucleoprotein. Thereafter, we used the BCA Protein Assay kit (Beijing Solarbio Science & Technology Co., Ltd., china) for quantifying protein content. After separating proteins through SDS-PAGE, they were transferred on the PVDF membranes. Then, each membrane was subjected to overnight incubation using primary monoclonal antibodies, including mouse anti-VEGFR2, rabbit anti-plexinA1 and mouse anti-gapdh (all Abcam, Cambridge, MA, UK), rabbit anti-Histon H3, rabbit anti-JAK, rabbit anti-STAT3, rabbit anti-p-JAK and rabbit anti-p-STAT3 (all ABclonal Co., Ltd,Wuhan, China) as well as mouse anti-β-actin (Cell Signaling Technology, Danvers, MA, USA) antibodies under 4°C. Afterwards, HRP-labeled secondary antibodies (Santa Cruz, CA, USA) were further used to incubate the membranes for 2 h. In addition, the enhanced chemiluminescence (ECL) substrate (Pierce, Rockford, IL, USA) was used for detecting protein blots, whereas the Image analysis software was employed for visualization.

Blood samples from cancer patients were collected for 15 min of centrifugation at 3000 ×g under 4°C. Human serum catecholamine and adrenaline, as well as VEGF extracellular of HUVEC levels were examined using the corresponding human ELISA kit (Mlbio biotech company, Shanghai, China) in accordance with specific protocols. In brief, the microplate reader was used to measure absorbance 450 nm.

In this experiment, the CORING Transwell chambers NO:3422 (Tewksbury, MA, UK) were used to quantify HUVEC migration. 200μl HUVEC (2×10⁵cells/ml) was seeded with serum free medium in the upper chambers, and chemoattractant in the bottom layers was RPMI-1640 that contained 10% FBS. After being cultured under 37°C for 24 h, we utilized cotton swabs for removing cells on surface of the upper chamber, whereas those penetrating the membrane were subjected to fixation and crystal violet staining. We selected 5 fields of view (FOV) to calculate the membrane-penetrating cell number.

The Matrigel Basement Membrane Matrix with high concentration was provided by BD Biosciences (Bedford, MA, UK). Add 50μl extracellular matrix gel to each well of pre-chilled 96-well plates, incubated in 37°C for at least 30min. HUVECs suspension (2×10⁴ cells/ml) were seeded onto the solidified gel for 12h. The inverted microscope was used to examine the endothelial tube structures, while Image-Pro Plus software was employed to quantify tubing area.

The BALB/c nude mice (male, aged 6 weeks, 16–20 g) were provided by Charles River Laboratories (Beijing, China) and raised within the individually ventilated caging systems. 200 μl MGC803 gastric cancer cells (1×10⁶cells/ml) were resuspended, then the mice were given subcutaneous injection of suspension in posterior region. Mice were given intraperitoneal injection with 5mg/kg/day ISO or 0.2 mg/kg/day ICI. The same amount of normal saline was injected to controls. Mice were sacrificed after 4 weeks.

Mice of subcutaneous tumorigenesis were anesthetized with 2% isoflurane dissolved within 100% oxygen at 1 L/min until they are unresponsive to external stimuli. Then mice were put onto the heated surface at 37°C with a continuous flow of isoflurane to ensure blood perfusion will not markedly be affected by temperature changes. Each mouse was placed onto the non-reflective light-absorbing plat in the prone position. Then, the laser Doppler imager together with acquisition software was initiated.

All assays were carried out in triplicates. Results were expressed in the manner of means ± SD. Student’s t-test and One-way ANOVA were utilized for statistically analyzing data. A difference of P<0.05 indicated statistical significance, as marked with asterisks in all figures.

A total of 332 (99.1%) patients completed the questionnaire. Among them, 24.7% of the cancer patients scored to be described as anxious, 57.5% were depressed according to those criteria. When combined into an anxiety-depression composite, the prevalence was 18.7%. There were 205 (61.7%) patients with negative emotions of anxiety or depression among the total sample size. The clinicopathological diagnosis and self-evaluation scores of the 215 tumor patients were shown as Table S1 . The SAS and SDS index of these patients were statistically analyzed according to different types of tumor. We found that the top three tumors with average SAS index were gastric cancer, renal cancer and tonsillar malignancy, ( Figure 1A ) and the top three tumors with average SDS index were tonsillar malignancy, glioblastoma and gastric cancer. ( Figure 1B ) The results showed that the anxiety and depression index of gastric cancer patients was the top three in all the cancers we detected.

According to index scores of ZSAS and ZSDS, these cancer patients were categorized into non-anxiety, anxiety and non-depression, depression group. The levels of chronic stress-related hormones in peripheral blood were measured. As a result, catecholamine contents markedly elevated among patients with anxiety and depression. ( Figures 1C, D ) Similarly, adrenaline had been found to increase significantly in patients with anxiety and depression. ( Figures 1E, F )

To examine whether chronic stress affects in vivo tumour formation, MGC803 cells were inoculated to mice subcutaneously. ( Figure 2A ) However, xenograft volume is not significantly different in ISO and ICI group. ( Figure 2C ) This finding consistent with a previous study by Caroline P. Le1, which suggested no influence of chronic stress on the growth of primary tumour (3). For assessing how stress affected the angiogenesis in tumour, we used Laser Doppler Blood Perfusion to quantify the levels of blood perfusion ( Figure 2B ). Then data of the recorded ROI perfusion values were evaluated by the software, and the mean of blood perfusion values was showed in Figure 2D . Collectively, these findings indicated isoprenaline acting on angiogenesis by effecting intratumoral blood perfusion values.

Angiogenesis requires vascular endothelial cells proliferation, migration and tube formation. Therefore, we treated HUVECs with 5, 10 uM ISO or 10 uM ICI for 6, 12, 24 hours. Isoprenaline markedly increased the ability of proliferation in HUVECs. In contrast, incubated with ICI reduced the ability of proliferation. ( Figure 2E ) Transwell assay showed that treatment of 5 uM isoprenaline promoted the migration of HUVECs, which was suppressed by ICI. ( Figures 2F, G ) Isoprenaline at growing concentrations and time was used to incubate HUVECs, in addition, there is another ICI incubated group. As suggested by these findings, isoprenaline promoted HUVECs tube formation depending on its concentration ( Figure 2H ) and treatment time ( Figure 2J ), whereas the increase was reversed by ICI incubated. The area statistics of tube formation of HUVECs are shown in Figures 2I–K . Collectively, these findings indicated isoprenaline acting on angiogenesis by effecting vascular endothelial cells proliferation, migration and the tube formation of HUVECs, which can be reversed by ICI.

To further detect what is the mechanism of stress on tumour angiogenesis, we assessed VEGF expression in HUVECs after incubated with ISO. HUVECs displayed markedly increased VEGF content following ISO treatment ( Figure 3A ). This suggests that stress can promote the VEGF autocrine of vascular endothelial cells. Then we used human recombinant VEGF165 to incubate HUVECs to detect changes in downstream molecules. Exposure to VEGF165 increased the VEGFR2 ( Figure 3B ) and plexinA1 ( Figure 3C ) mRNA level within HUVECs depending on its concentration. Also, VEGF165 treatment increase protein levels of VEGFR2 and plexinA1. ( Figure 3D )

As suggested by these findings, ISO treatment significantly promoted the VEGF autocrine, and VEGF further increases the expression of VEGFR2 and plexinA1. To further confirm this process, we use ISO and ICI, the agonist and the specific blockers of β2-AR to incubate HUVECs respectively. The β2-AR expression were obtained from western-blotting analysis. ( Figure 3F ) Further, our findings revealed that VEGFR2 and plexinA1 upregulated by ISO stimulation, whereas ICI treatment reduced mRNA ( Figure 3E ) and proteins ( Figure 3F ) expression of VEGFR2 and plexinA1. Cabozantinib, a selective inhibitor of VEGFR2 was selected in this study to elucidate the role of VEGFR2 and plexinA1 in ISO promote angiogenesis. Study has shown the IC50 of cabozantinib almost >400 nM (18). Hence, we selected 40 nM cabozantinib as research concentrations in HUVEC and hCMEC/D3. Our results indicated that blocking VEGFR2, ISO act on vascular endothelial cells through plexinA1. ( Figure 3G ) Collectively, our observations showed that ISO/β2-AR-VEGF-VEGFR2/plexinA1 signaling served as a regulator in mediating the tumour angiogenesis.

To further verify the function of plexinA1, we introduced the shplexinA1 interference sequence or control shNC in HUVECs. The transfection efficiency was approximately 76% based on the percentage of GFP-positive cells. ( Figure 3I ) Then, CCK8 and transwell assay was applied in combination with tube formation assay for evaluating the impact of plexinA1 on HUVECs by interfering plexinA1 expression. We found shplexinA1 can significantly decrease the ability of proliferation ( Figure 3H ), migration ( Figures 3J, K ) and tube formation ( Figures 3L, M ) in HUVECs. Collectively, our findings indicated plexinA1 is indispensable in tumour angiogenesis.

We found that ISO not only promoted the expression of plexinA1 and VEGFR2, but also activated the JAK2-STAT3 pathway. ISO mainly promoted the phosphorylation of JAK2 and STAT3, and increased the translocation of p-STAT3. ( Figures 4A–C ) To further detect the mechanism of between plexinA1 and upstream and downstream signals, we tested the effects of shplexinA1 on JAK2-STAT3 pathway by western blots. Inhibition of plexinA1 strongly decreased the expression of VEGFR2, and phosphorylation of JAK2 and STAT3. At the same time, inhibition of plexinA1 also decreased the translocation of p-STAT3. ( Figures 4D–F ) These results suggested that ISO-plexinA1/VEGFR2-JAK2-STAT3 signaling may be related with tumour angiogenesis. In order to verify the hypothesis, AG-490, a phosphorylation inhibitors of JAK2, as well as stattic, a phosphorylation inhibitors of STAT3 were used respectively to detect that isoprenaline promote activation of HUVECs. PlexinA1 and VEGFR2 protein expression was not affected by JAK2-STAT3 pathway inhibitors, but could be activated by isoprenaline. ( Figure 4G ) Isoprenaline had no effect on the expressions of JAK2 and STAT3, but increased the expressions of p-JAK2, p-JAK3 and the translocation of p-STAT3 in HUVECs. ( Figures 4G, I, J ) Similar results were observed in hCMEC/D3 cells at protein levels. ( Figure 4H ) These results indicated that isoprenaline could promote plexinA1 and VEGFR2 protein expression, and then activate JAK2-STAT3 pathway.

All together, these results suggested that chronic stress promote plexinA1/VEGFR2-JAK2-STAT3 in VECs to promote angiogenesis.