The research design and experimental technology were reviewed approved by the Institutional Review Committee, which is the Institutional Animal Care and Use Committee of Guangdong Pharmaceutical University. Purchase of SPF nude mice at Guangdong Animal Center (Certificate No. SYXK (Guangdong) 2017-0125). All animal handling and procedures follow the international guidelines for the use and care of laboratory research animals and comply with the regulations of the Institutional Animal Care. HCT116 cells were purchased from ATCC. Provided with technical support from Saiye Biological Co, Ltd, the HIF1α gene of HCT116 cell line was knocked out using CRISPR/Cas9 technology to obtain HIF1α (-/-) cell line, HCT116 cells were established by virus transduction technology to establish a stable transfected strain overexpressing HIF1α gene, and the obtained HIF1α OE cell line. The D-lactic acid detection kit was purchased from AmyJet Scientific Co.Ltd. Standards for putrescine, spermidine and spermidine were purchased from Pfifizer. Chromatographic grade methanol and acetonitrile were purchased from Opple. BCA kits were purchased from Thermo. PVDF membranes were purchased from Bio-Rad, USA. Primary antibodies for ODC, c-Myc and HIF1α were purchased from Abcam. Primary antibodies for ZO1, Occludin, spermidine/spermine N1 acetyltransferase (SSAT), OAZ1 and β-actin, and rabbit, mouse secondary antibodies were purchased from Proteintech. The remaining reagents were purchased from Guangzhou Chemical Reagent Factory.

HCT116 cells with different expression levels of HIF1α were inoculated subcutaneously. When the tumor tissues were 10 mm × 10 mm, they were transplanted into the armpits of nude mice. According to the expression level of HIF1α, nude mice were divided into normal group, HIF1α (+/+) group, HIF1α OE group, HIF1α (-/-) group. Tumor-bearing mice with different expression levels of HIF1α were reared for four weeks and their body weights were recorded every other day. The body weight of tumor-bearing mice was found to decrease slightly or remain stable from around day 14, indicating that the intestinal mucosa of nude mice may be damaged at this time. The tumor-bearing mice in the HIF1α (+/+) group were gavaged spermidine (20mg/kg) for two weeks after inoculation for two weeks. During the administration period, the mental state, vital signs and food intake of nude mice were observed and recorded every day, and the tumor volume was observed and recorded every other day. After two weeks of gavage, the nude mice were sacrificed after 8 hours of fasting tumors were stripped and weighed. The small intestines of nude mice were collected and the intestinal mucosa was scraped. The tumor volume was calculated by the metric: V=0.5×a×b² (a: tumor length; b: short tumor diameter).

Blood was collected from the eyeballs of nude mice in each group, and the blood was allowed to stand at room temperature for 30 min, and the supernatant was aspirated after low-temperature centrifugation. Serum D-lactic acid concentration of nude mice in each group was determined by Ai Meijie serum D-lactic acid kit. A standard curve was drawn based on the reaction between the D-lactic acid standard solution and the enzyme. In a 96-well plate, the prepared D-lactic acid working solution and serum samples were added, and incubated at room temperature for 1.5 h in the dark. The absorbance at 575nm and 605nm was measured by a microplate reader, and the ratio of OD575nm/OD605nm was calculated. Substitute into the standard curve to obtain the D-lactic acid concentration in the sample.

When dissecting the tumor bearing mice, part of the tumor tissue and intestinal mucosal tissue were fixed in 4% paraformaldehyde for 24 hours, the fixed solution was washed off, and then the tissues were dehydrated in different concentrations of ethanol. Put the dehydrated tissue into xylene for transparency and paraffin with different melting points. Slice the embedded wax block and bake it in a 60°C constant temperature oven for 2h. Dehydrate and transparentize again, and finally drop an appropriate amount of neutral gum in the center of the slide, the morphological changes of different groups of tissues were observed by optical electron microscope.

(1) Put the tissue into a grinder for full grinding, quantitatively detect the protein content through BCA, adding 5% perchloric acid of 5 times the volume to the rest of the mixture, and take the supernatant after centrifugation for derivatization. (2) Add NaOH and benzoyl chloride to the tissue supernatant, add chloroform (shading) after full reaction, suck the lower organic layer after centrifugation, add ultra pure water, suck the lower solution after centrifugation, blow dry the extract with nitrogen, add 1mL acetonitrile to dissolve the residue and filter. (3) Chromatographic conditions: 20 μL for each loading, the chromatographic column is C18 column, the detection wavelength is at 229 nm, the mobile phase is acetonitrile and ultra pure water (38:62), and the flow rate is 1 ml/min,

Determine the protein concentration of the sample by BCA method, prepare a gel of appropriate concentration, inject the protein sample into the gel hole, after electrophoresis, transfer the protein to PVDF membrane, block it with 5%BSA, incubate it with primary antibody at 4° covernight, incubate it with secondary antibody at room temperature for 1 hour the next day, wash the PVDF membrane with TBST, adding luminescent solution and put it into the imager for development, and analyze the gray value with software.

HCT116 cell suspensions with different expression levels of HIF1α were injected subcutaneously into the armpits of nude mice to construct xenograft models. The expression of HIF1α protein in the tumors of HIF1 (+/+) group, HIF1α OE group and HIFα (-/-) group was verified by Western blot ( Figure 2A ), and the specific expression amount ( Figure 2B ). Compared with HIF1α (+/+) group, the expression of HIF1α protein in tumor tissue of HIF1α OE group was significantly increased. Although the HIF1α protein expression was significantly decreased in the tumor tissue of the HIF1α (-/-) group, there was still a small amount of expression. This indicates that the tumor-bearing mouse model was successful. During the rearing period, tumor major and minor diameters were measured by vernier calipers every other day. Compared with HIF1α (+/+) group, HIF1α (-/-) group had the slowest increase in tumor volume and HIF1α OE group had the fastest increase in tumor volume ( Figure 2C ). Four weeks later, tumors in the armpits of nude mice were dissected out and photographed ( Figure 2D ). As exhibited in Figure 2E , the weights of nude mice with different expression levels of HIF1α was lower than that of the normal group. The weight of HIF1α OE group is lower than that of the HIF1α (+/+) group, and the weight of the HIF1α (+/+) group is lower than that of the HIF1α (-/-) group. As shown in Figure 2F , the tumor mass of HIF1α OE group was larger than that of HIF1α (+/+) group and HIF1α (-/-) group. The above results suggest that the activation of the HIF1α pathway can promote tumor growth.

The differences in intestinal mucosal permeability of nude mice in each group are shown in Figure 3 . When the intestinal mucosal barrier is damaged, D-lactic acid, one of the products of bacterial growth and metabolism in the gut, enters the circulating blood through the intestinal mucosal tissue. The higher the D-lactic acid concentration, the greater the permeability. The serum D-lactic acid concentration in the normal group was the lowest, and the serum D-lactic acid concentration in the tumor-bearing mice groups were higher than that in the normal group. The HIF1α OE group was higher than that of HIF1α (+/+) group, and that of HIF1α (+/+) group was higher than that of HIF1α (-/-) group. This indicates that the intestinal mucosa of the HIF1α OE group had the most severe intestinal mucosal damage and the greatest permeability ( Figure 3A ). Western blot was used to detect the expressions of tight junction proteins: ZO-1, Occludin and Claudin ( Figure 3B ). As shown in Figure 3C , the results showed that the tight junction proteins in the intestinal mucosa of tumor-bearing mice with three different HIF1α expression levels decreased in different degrees, and the degree of decline was: HIF1α OE group > HIF1α (+/+) group > HIF1α (-/-) group. Because the tight junction proteins are an important part of the first line of defense of the intestinal mucosal barrier, so this result indicates that with the up-regulation of HIF1α, the intestinal mucosal damage is progressively worsened.

Observed pathological sections under a 100x optical microscope. In the normal group, the intestinal mucosal glands were neatly arranged, erect, and close to the muscularis mucosae ( Figure 4A ). In the wild-type HIF1α (+/+) group, the villi were obviously broken, the crypt depth became deeper, and part of the central chyle duct disappeared. As shown in Figures 4B–D , Compared with the HIF1α (+/+) group, the glandular arrangement and integrity of HIF1α (-/-) group were improved, and the length of villi were increased; the glandular structure of HIF1α OE group was mostly disappeared, and the muscularis mucosa was thinned, the villi were shortened and the cells were disorganized. This indicates that the intestinal mucosal structural damage in tumor-bearing mice is aggravated after the activation of the HIF1 pathway. As shown in Figures 4E–G , observed under 200x microscope, tumor cells in HIF1α (+/+) group showed obvious atypia, with spindle-shaped tumor cells, rounded nuclei, and increased nuclear division. Compared with the HIF1α (+/+) group, the tumor parenchyma in the HIF1α (-/-) group was reduced. In the HIF1α OE group, the tumor parenchyma increased, the cell structure was blurred, the border was unclear, the nucleus was not obvious, and there were some tumor cells showed sheet necrosis.This may be because the tumor is too large, causing internal ischemia, hypoxia and necrosis.This may be caused by internal hypoxia-ischemia due to the large tumor.

The expressions of glycolysis-related enzymes (HK2, PKM2, PDK1 and LDHA) in HIF1α (-/-) group, HIF1α (+/+) group and HIF1α OE group were detected by Western blot ( Figure 5A ), and as shown in Figure 5B , the analysis showed that the gray value increased with the increase of HIF1α expression. These results indicate that the up-regulation of HIF1α expression can promote the glycolysis process of tumors. Since tumors obtain energy through glycolysis, up-regulation of HIF1α expression promotes tumor growth, which is also consistent with the results in 3.1.

In this paper, high-performance liquid chromatography was used to detect the content of polyamines in tissues. As exhibited in Figure 5C , the content of polyamines in intestinal mucosa of normal group was significantly higher than that of HIF1α(+/+) group. Compared with the HIF1α (+/+) group, the polyamine content in the HIF1α (-/-) group was increased, while the content in the HIF1α OE group was decreased. However, as showed in Figure 5D , the levels of polyamines in tumor tissue were opposite to those in intestinal mucosa: HIF1α OE group > HIF1α (+/+) group > HIF1α (-/-) group. The content of polyamines in tumor tissues with high expression of hif1 is high, and the content in intestinal mucosa tissue is low, while the expression of hif1 silence is the opposite. This indicates that the tumor preys on intestinal mucosal polyamines. As shown in Figure 5E , WB was further used to detect the expression levels of polyamine transporter SLC22A and polyamine synthesis rate-limiting enzyme ODC. The results of gray value analysis showed that with the up-regulation of HIF1α expression, the protein expressions of SLC22A increased ( Figure 5F ). SLC22A as a polyamine transfer protein, its increased expression suggests that overexpression of HIF1 promotes tumor transfer polyamines from outside. ODC is the rate-limiting enzyme for polyamine synthesis. The expressions of c-Myc and ODC decreased ( Figures 5A, F ), and the degree of decline is HIF1α OE group > HIF1α (+/+) group > HIF1α (-/-) group. This indicated that overexpression of HIF1α could inhibit the transcription of c-Myc target gene ODC. Decreased ODC reduces endogenous polyamine synthesis thereby promotes the predation of polyamines in the intestinal mucosa by the tumor to meet the tumor’s own growth needs.

During the gavage period, the long and short diameters of the tumors were measured using vernier calipers every other day. Increased tumor volume in the spermidine-administered HIF1α (+/+) group compared to the non-administered HIF1α (+/+) group ( Figure 6A ). After two weeks of gavage, it was found that the tumors in the spermidine group were larger ( Figure 6B ). As exhibited in Figure 6C , the weight of spermidine group was higher than that of the HIF1α (+/+) group after administration. As shown in Figure 6D , After the tumors were weighed, it was found that the tumor mass in the spermidine group was higher than that in the HIF1α (+/+) group. The results of the intestinal mucosal permeability test after spermidine administration are shown in Figure 6E . The serum D-lactic acid concentration in the HIF1α (+/+) group was higher than that in the spermidine group and higher than that in the normal group. Using WB method to detect the tight junction proteins in the HIF1α (+/+) group and the spermidine group ( Figure 6F ), we found that the protein expressions of ZO1 and Occludin were significantly up-regulated in the spermidine group ( Figure 6G ). The results of histopathological sections can be seen in Figure 7 . After supplementation with exogenous polyamines, compared with the HIF1α (+/+) group, the spermidine group had significantly increased intestinal glands, closely arranged, and increased Paneth cells And the blood vessels in the tumor tissue increased significantly. Increased blood vessels is an important condition for tumor growth, infiltration and metastasis, and it is more conducive for tumors to absorb nutrients from the body for their own growth. Apparently, supplementation with exogenous polyamines protects the intestinal mucosa and promotes tumor growth in tumor-bearing mice. As shown in Figures 6H, I , the content of polyamines in the intestinal mucosa tissue of the normal group is the highest. The content of polyamines in intestinal mucosa tissue and tumor tissue in spermidine group was higher than that in HIF1α (+/+) group. Western blot was used to detect the expression levels of glycolysis-related proteins and SLC22A protein ( Figures 6J, K ), and the gray value analysis displayed that the expressions of these proteins were all up-regulated ( Figure 6L ). This indicates that supplementation of exogenous polyamines can promote tumor glycolysis, meet the needs of the tumor itself and avoid the predation of polyamines by the tumor on the intestinal mucosa.