Can Serum Thymidine Kinase 1 Detect Small Invisible Malignant Tumours?

Objectives. Early detection of malignant tumour is a prerequisite for a successful treatment. Here we investigate if thymidine kinase 1 is more sensitive than imaging technology to discover small invisible malignant tumours. Material and Methods. The cellular concentration of TK1 was determined by a novel automatic chemiluminescence analyzer of magnetic particle immune sandwich minimum. The primary and secondary antibodies linked to the magnetic beads were chicken anti-human thymidine kinase 1 IgY-polyclonal antibodies (IgY pAb). The minimum number of cells able to be detected by the novel detection technology using an automatic chemiluminescence analyzer were determined based on the cellar TK1 concentration of low and high TK1 cell lines of known cell count. Results. The TK1 concentration of malignant cell was found to be 0.021 pg/cell. Assuming 200 pg of total protein/cell, TK1 corresponds to 0.01 % of the total protein/cell. The concentration of TK1 in human blood serum of malignant patients is in the range of 2-10 pmol/l (pM), corresponding to about 50 x10 6 growing cells in the body that release TK1 into 5 litre blood. The limit visibility by imaging of a tumour is about 1 mm in diameter, corresponding to about 10 9 cells of a cell diameter of 1µm. Conclusion. TK1 is more sensitive than imaging.


Introduction
Early detection of malignant tumour is a prerequisite for successful treatments. Therefore, imaging technologies have been developed during the last decades. However, the imaging methods so far is still not sensitive enough to discover very small tumours. In order to solve this limitation, extensive research has been performed to found tumour-related biomarkers in serum sensitive enough to discover small tumours. Thymidine kinase 1 (TK1) is such a tumour-related biomarker. TK1concentration in serum (STK1p) can effectively discover persons with high-risk precancerous diseases that progress to malignant tumours. 1,2 STK1p can also be used to monitor the curative effect, prognosis and risk of recurrence of patients with malignant tumours. 1,2 In addition, TK1 in serum discover tumour in the body before discovered by imaging. In patients with breast and colorectal malignancies, it was observed that TK activity in serum was elevated already nine months before manifested by imaging. 3 TK1 is a kinase enzyme that is expressed in the cytoplasm of the cell and converts deoxythymidine (dThd) into deoxythymidine monophosphate (dTMP), a key enzyme introducing dThd into DNA through the salvage pathway, and thus, closely related to cell proliferation.
According to our original STK1p assay, an indirect ELISA assay adaptived enhancing chemiluminescent immune-dot blot biotin detection system (ECL dot blot), STK1p value appear earlier than tumour visible by imaging. 1,2 So far, there is still lacking quantitative data on the number of tumour cells in the body needed to release detectable TK1 into serum and also in relation to the size of tumour by imaging. To be able to correlate the STK1p concentration to the number of tumour cells in the body that released it, we need to used cell extract of tumor cells growing in culture medium (in vitro), since the ECL dot-blot assay gave non-immune cross-reactivity. Fortunately we found that the sandwich ELISA assay adaptived automatic chemiluminescence biotin system based on magnetic beads did not give cross reactivity when determine TK1 in cell extract.
Here we set up a novel automatic chemiluminescense biotin system to be able to determine the correlation between TK1 value and cell number in cell lysates of proliferatsng tumor cells. This is the rst study try to determine the number of cells in the body that releases detectable STK1p. We con rmed that the serum thymidine kinase 1 can detect small invisible malignant tumours.

Study design
In the rst part, we investigated whether the automatic chemiluminescence magnetic bead platform is able to determine TK1 in a reliable way. In the second part, we correlate the TK1 concentration in serum (STK1p) to the number of cells that generate detectable concentration of TK1 in serum. In the third part, we show data from health screening in which elevated STK1p correlate to the risk to develop premalignancy and malignancy later in life, before any indication of presence of tumours by imaging. Finally, the results are summary schematic showing the relation between TK1 in serum, imaging of tumours and tumour growth.

Cell cycle analysis
TK1 in relation to cell cycle was done in HeLa cells by centrifugation elutriation method as described previously. 4 The proportion of G 1 , S-phase and G 2 + M was done by ow cytometric analysis described previously. 4

Serum samples
Serum samples used in this study had no symptoms of tumour-related diseases or no infectious diseases, as assessed by health examination. Some persons might have mild in ammation, or chronic proliferating disease. All samples showed low STK1p values. [5][6][7][8] All participants provided informed consent before entering the study, which was conducted in accordance with the Declaration of the 1964 Helsinki Declaration and the Harmonized Tripartite Guideline for Good Clinical Practice from the International Conference on Harmonization. The study shown in Fig. 2  2.6. Automatic chemiluminescence biotin system assay of TK1 The novel detection technology based on an automatic chemiluminescence analyzer was set up for detection of small invisible malignant tumours. The novel technical principle of the system is magnetic particles linked to the primary TK1 IgY polyclonal antibody (pAb) and the secondary biotin TK1 pAb and the test sample or the calibrator to form a sandwich complex. Subsequently, alkaline phosphataselabelled streptavidin reagent is added, and streptavidin speci cally binds to biotin and plays a signal ampli cation effect.
The test procedure is as follows:

Preparation of magnetic particles
The preparation of the magnetic beads was done according to the company's manual (KeySmart Ltd., China), brie y: 5mg 50µl MS300 Tosyl magnetic beads was add in a magnetic eld, removing the supernatant and then adding 2-20 times activation buffer, shaking and washing the magnetic beads.
After that, the supernatant was removed and washed again twice for 10 mins.

Couple of the primary human IgY pAb to the magnetic particles
The primary human IgY pAb was coupled to the magnetic beads by adding at a ratio of 1000:1-10 to a magnetic bead in catalyst and mixing solutions, and then incubated for 18 hours at 37°C. The coupling reaction was blocked by adding 10% BSA and continue to react for six hours at 37°C. The supernatant was removed by put the mixture in a magnetic eld, washing the magnetic TK1 beads with a cleaning solution at 4 times, dilute to 1mg/ml, and stored at 2-8°C.

The formation of the sandwich complex
Adding 10 µl of cell lysates and TK1 calibrator, respectively, for 10 minutes at 37°C, washing x 3 times. And then added 10 µl of the biotin-human TK1 IgY pAb for 10 minutes at 37°C, and washing x 3 times.
2.6.4 Immune response signal by alkaline phosphataselabelled streptavidin reagent Added 100µl of alkaline phosphatase reagent labelled with streptavidin for 10 minutes at 37°C, washing x 3 times, and then adding 100-µl of the luminescent substrate, reacting for 5 minutes. The content of TK1 in the sample was positively correlated with the luminescence value, and the content of TK1 was calculated from the luminescence standard curve.

Antibodies
We used chicken anti-human TK1 IgY polyclonal antibody (HTK1-IgY pAb) raised against a peptide (a residue of 195-225, GQPAG PDNKE NCPVP GKPGE AVAAR KLFAPQ synthesized by this company Hybio Pharmaceutical Co., Ltd, China). 9 Two different batches of the antibody were used for collection and detection of TK1, respectively, in an automatic analyser (Automatic Chemiluminescence Immunoassay Analyser, Keysmile Smart 6500h, China).

ECL dot blot immune-detection system
A commercial ECL dot blot immune-detection kit was used to determine serum thymidine kinase 1 concentration (STK1p) (SSTK Inc., Shenzhen, P.R. China, www.sstkbiotech.com). Serum samples were probed with the anti-chicken HHTK1-IgY pAb, the same HTK1 IgY pAb as used in the automatic chemiluminescence assay. Brie y, three µl of serum were directly applied onto a nitrocellulose membrane (HybandTM-C, Amersham). TK1 calibrators (2.2, 6.6 and 20 pmol/L, SSTK Inc., Shenzhen, P.R. China, www.sstkbiotech.com) were used. The membrane was blocked in TBS (Tris-buffered saline) with 6% nonfat milk for 30 mins and incubated at room temperature for 1 h after addition of the primary biotinylated HTK1-IgYpAb. Then the membrane was incubated in TBS buffer with Streptavidin Horse-Radish-Peroxidase (SA-HRP), followed by addition of Enhanced Chemiluminescence (ECL) substrate. The light intensity of a single spot on the membrane was detected using a CIS-II Imaging System (SSTK Ltd, Shenzhen, China). Based on the light intensities of known concentrations of calibrators, the light intensities of the TK1 spots were re-calculated and expressed as pmol/L (pM). All experiments were performed in a blinded manner and in duplicate. 5 washed twice. Cells were then re-suspended to a nal concentration of 5×10 6 cells/ml in a lysing buffer (10 mM Tris buffer, pH 7.5, 0.5% P-40 (V/V), 2 mM β-mercaptoethanol, 5 mM NaF, 5 mM MgCl 2) . The cytoplasmic fraction containing TK1 was obtained by centrifugation at 4,000 g for 5 mins and further centrifuged at 48,000 g at 5 mins. Fifty micro-litres of the supernatant were added to 200 µl of a cocktail (55 mM Tris-HCl, pH 7.5, 5.6 mM NaF, 5.0 mM ATP, 3.8 mM MgCl 2 , and 0.06 mM unlabelled dThd and 2.3 µCi 3 H-dThd [speci c activity 0.74 TBq or 20 Ci/mmol, New England Nuclear, Boston, MA) and incubated for 60 mins at 37˚C. The mixture was applied to DE-81 paper disks (Whatman) and washed. The radioactivity in the paper disk was measured in a beta-counter. The TK activity was expressed as U/L. Since the TK activity is low in G 1 , but high in S + G 2 , the TK activity was also expressed as DPM/S + G 2 cells/min. The standard deviation was no more than 10%. 4

Statistical analysis
The statistical signi cance for the correlations between parameters was calculated by correlation-Pearson test (SPSS Statistics V25.0, IBM, USA). P-values of < 0.05 were considered as statistically signi cant.

The accuracy of the magnetic bead platform
The accuracy of the chemiluminescence magnetic bead platform was tested by a standard curve (Fig. 1), a recovery test (Table 1), a deviation test (CV) ( Table 2) and by distribution of TK1 in serum (Fig. 2).

Standard curve
In the standard curve (Fig. 1)

Recovery test
The recovery test was performed by determine TK1 concentration before and after adding a known concentration of TK1 calibrators (1.10pM and 8.00pM). The determination of TK1 calibrator was repeated three times. The recovery rates were 99.18% and 100.26%, respectively, within acceptable range (Table 1).

STK1p distribution in serum
The STK1p concentration in serum of normal healthy people is very low or almost impossible to detect using the original STK1p assay of ECL dot-blot. According to recent health screening data in China, the frequency of disease-free person is only 1-2%. [5][6][7][8] Here we collected serum sample from 115 persons who had no symptoms of tumour-related diseases, as assessed by health examination. The STK1p value was almost normal distributed in the range of 0.01-0.60 pM, with a tail up to 2.65 pM (Fig. 2).
The elevated STK1p values in serum above 0.60 is most likely to be of people with mild infections/in ammations or chronic proliferation diseases. [5][6][7][8] When compared the chemiluminescence magnetic bead platform assay with the original STK1p assay (ECL dot-blot) it was found that the chemiluminescence magnetic bead platform improved the sensitivity three-fold with a sensitivity down to 0.01pM of STK1p in serum. Thus, the chemiluminescence magnetic bead sandwich TK1 kit used in the automatic device is sensitive and speci c enough to be able to detect low concentration of TK1 in serum, corresponding to very small tumours.

TK1 in relation to cell cycle.
To prove that TK1 is a proliferation-biomarker, its relation to cell cycle was performed in HeLa cells. The cell cycle stages were isolated by centrifugated elutriation and determined by ow cytometry DNA measurements, as described previously. 4 TK1 activity and TK1 protein concentration were low in G 1 , elevated in S-phase and then decline in G 2 + M (Fig. 3).

TK1 in growing cells
The TK1 concentration in cells were determined by an automatic chemiluminescence biotin system assay based on magnetic beads in cell extract of both TK1 low (TK1 − ) and TK1 high (TK1 + ) cell lines, growing in culture medium. The TK1 value of the TK1 low-cells contained about 2% of TK1 compared to the TK1 high-cells, no non-speci c immune cross-response reaction was found (Fig. 4).
Dilution of the high-TK1 cell extract showed a linear decrease in the OD-values, indicates that the magnetic bead platform is able to measure TK1 of cell extract in a reliable manner. Based on the standard curve, the TK1 concentration per cell was calculated to be 0.021pg. Assuming a total protein concentration per cell of 200pg, a concentration often found in in vitro growing cells, TK1 corresponds to 0.01% of total protein concentration of a cell. This is a likely TK1 value and support that the TK1 automatic magnetic platform we use is reliable for measuring TK1 concentration.

Calculation of TK1 value in relation to cell number and imaging
Recently studies show that TK1 in serum of tumour patients is in the range of 0.1-20pmol/l, (pM). 1,2 In 5 litre blood, a normal blood volume in an adult person, 2pmol/l (pM) of TK1 corresponds to 1.1µg total TK1. Suppose 0.021pg TK1/cell, 1.1µg TK1 in 5liter blood corresponds to 52.4 x 10 6 growing tumour cells that release TK1 (total 1.1 µg divided by 0.021 pg/cell = 52.4 million cells). That should be compared to the number of cells needed to visible a tumour in imaging of at least 10 9 cells (calculated of a tumour of a diameter of 1mm, containing tumour cells with a diameter of 1µm).

TK1 in serum (STK1p) correlates to development of malignancy
In the health screening follow-up in this study, we found that elevated STK1p values increase the risk to develop premalignant and malignant tumours and death within 11 years (Fig. 5).
Randomly In the elevated STK1p group, 16 persons (9.4%) developed new malignancies (carcinoma of gastric, liver, ovary, prostate) within 132 months and ve persons (2.9%) died in their cancer diseases (carcinoma of gastric, liver and breast), within 132 months. Eleven persons (6.5%) showed progress in their premalignancy diseases up to 132 months (breast and prostate hyperplasia, HBV of high risk). Twenty-three persons did not show any further progress up to 132 months (data not shown). The tumours were discovered by B-ultrasound in the follow-up health screening study and con rmed by imaging.

Relationships between TK1 and tumour growth in serum
The relationships between TK1 and tumour growth is visualized in Fig. 6, based on data presented here and in previous studies. 1,2 The tumours are visible in imaging when the tumour volume/number of tumour cells excide about 1 mm/10 9 million cells. On the contrary, tumours are detected by TK1 in serum at a concentration of TK1 corresponding to about 50 x 10 6 cells. Hence, TK1 concentration in serum is more sensitive to detect tumours than imaging.
However, it should be noted that these correlations are only found at early and middle stages of tumour growth. In many cases, serum TK1 value decreases in later stages of tumour growth. When the tumour size increases, and be able to be detected by imaging, the growth rate decreases due to lack of nutrition and as a consequence, reducing the TK1 concentrations in serum.

Discussion
TK1 in serum (STK1p) was determined by an automatic magnetic bead analyzer based on chemiluminescence sandwich magnetic bead immune technology, with high sensitivity and accuracy. Now through this method, we veri ed that it is possible to detect the correlation between the increase in TK1 and the proliferating tumor cells.
Chicken anti-human TK1 IgY polyclonal antibodies recognized ve different epitopes on the C-terminal end of TK1 (1 GQPAGPDNKENCPVP; 2 GEAVAARKLF; 3 NCPVPGKPGEAV; 4 NCPVPGKPGE; 5 GQPAGPDKEN), where three epitopes (No 2, 4, 5) showed signi cantly higher a nities than the others. 1 This part of TK1 is responsible for the cell cycle regulation of TK1, and thus related to cell proliferation 12 .
The novel accuracy tests consist of magnetic particles coupled to the primary TK1-pAb which bind to the test sample or calibrator, while the secondary biotinylated-TK1 pAb bind to the complex TK1-pAb/sample/calibrator to form a sandwich complex, show high accuracy. It con rming that the automatic chemiluminescence magnetic bead biotin-immune platform is a reliable method determining TK1 in tumor cells and serum, no non-speci c immune cross-response reaction was found. The deviation was about 1%, which is speci city and signi cantly better than the original serum dot blot TK1 assay (deviation about 15%). The distribution of TK1 concentration in serum also shows that the automatic chemiluminescence immune magnetic bead biotin-immune platform is sensitive down to 0.01pM. This is crucial, since the TK1 concentration in serum is very low (< 0.10pmol/l (pM)).
Based on the cellular concentration of TK1 (0.021pg/cell), we were able to translate the TK1 concentration in the serum to the number of cells in the body that releases TK1 into the serum. Two pmol/l (pM) TK1 in serum corresponds to about 50 x 10 6 cells in the body that release TK1 into serum, which is about 20 times lower than needed to visualize a tumour by imaging (about 1 mm/10 9 cells).
Thus, TK1 in serum is more sensitive than imaging.
In addition, the anti-human pAbs have the bene t over anti-human TK1 IgG antibodies because of the interfering with human anti-mouse IgG monoclonal antibody (HAMA), 13 which often causes antibody response effects. In terms of tumour treatment, especially when mouse monoclonal antibodies are used for treatment, the use of monoclonal IgG antibody type, tumour-related markers for detection may increase the incidence of HAMA in the body. The advantage of the chicken antibody detection system to monitor the e cacy of treatment is that it does not react with HAMA. Therefore, immunoassay method using chicken antibodies should theoretically be more reliable than using mammalian antibodies.
Furthermore, since there is both genetic and species differences between chicken IgY and human IgG, the risk of unspeci c immune reaction is low using chicken IgY antibodies. In addition, IgY antibody does not activate the human complement system, or show Rheumatoid reactions, thereby partially blocking the activation of non-speci c antigen binding sites in human serum. Compared with polyclonal antibodies prepared by traditional rabbit immunization, chicken IgY has endogenous molecular homogeneity (only one type of antibody molecule is produced, namely IgY). Finely, polyclonal antibodies in general show more advantages than single antibody detection because of binding to higher number of epitopes, giving more accurate results.
It is known from recent studies that TK1 is closely related to the cell cycle, and thus related to proliferation. 1,2 The TK1 antibodies we developed and used in this study con rmed that. The level of TK1 starts to increase at the boarder of G 1 /S phases of the cell cycle with a maximum value during S-phase and gradually decreases as the cell enters late G 2 -phase/mitosis. TK1 is not a speci c tumour-related marker, but a proliferation marker. That means that TK1 is useful both in non-malignant and malignant growing cells. To be able to assess if the elevated TK1 value in serum indicates presence of tumours, we strongly recommended to contact an oncology clinic for further investigation. TK1 in health screening should be regarded as the rst warning of the presence of tumours or the risk to develop malignancy later in life. When using TK1 of persons already diagnose for malignancy, TK1 value in serum or in the tumour tissue (immunohistochemistry) indicates the proliferation degree of the tumour and is a reliable value of the prognosis. Our screening study shows that TK1 in serum not only warning for the presence of tumours in the body, but also for the risk to develop malignancies and recurrence and the risk to dye (survival).  STK1p distribution determined by a chemiluminescence magnetic bead platform assay (Keysmile) and an ECL dot blot immune-detection system. The serum samples (n=115) were collected in a health screening study in Fujin, China, of a cohort of 160,000 people.

Figure 3
Concentration and activity of TK1-and TK1+ cell lines. A) The concentration of TK1 was determined by Western blot and B) the protein was determined by Western blot shown in A) and the activity by a TK1 isotope assay. The cell cycle composition was analysed by DNA ow cytometry.   A schematic diagram of correlation between human TK1 concentration in serum and tumour growth.
Clinical stages refer to visible tumours by imaging.