br the calculated value of max Kd and

(2), the calculated value of λmax, Kd, and n is 718.9 pm, 64.0 ng/mL and 0.10, respectively. The Hill coefficient n is less than 1, indicating the negative cooperativity with respect to ligand binding to the re-ceptor. The molecular weight of NSE is approximately 47 kDa, resulting in a dissociation constant Kd of 1.36 × 10−9 M which agrees with the previous research (Acero Sánchez et al., 2016). Based on the experi-mental results, the limit of detection (LOD, at a signal/noise ratio of 3) of NSE concentration is calculated below 1.0 pg/mL, which is 4 TRIzol of magnitude lower than NSE cut-off value (15.2 ng/mL) of SCLC (Wang et al., 2013), demonstrating the possibility for quantitative detection of NSE in clinical samples. The enhanced biosensing sensitivity is 100-fold higher than GO- and AuNPs-based biosensors (Zhou et al., 2009; Qu et al., 2011). Due to its ultrahigh surface-to-volume ratio, wide tunable range of bandgap and superior molecular adsorption energy, BP plays an important role as bio-nano-photonic interface which increases the sensing area, number of binding sites, and affinity binding efficiency, hence amplifies the optical signal providing superior biosensing per-formance.
For the practical biosensing application, the reusability is an im-portant function. The reusability of the proposed BP-TFG has been 
examined by detecting the binding interaction in 10 ng/mL NSE for multiple times. Fig. S5 presents the comparison results for three cycles with the percentages of resonant wavelength shift and initial binding rate. The maximum wavelength shift as the absolute RI change after reducing the baseline signal in PBS prewashing stage retained 95.5% and 91.5% after second and third cycle, respectively. Likewise, the in-itial binding rate calculated with the data over the first 3 min of binding interaction maintained greater than 89% and 88.5% after second and third cycle, respectively. These results confirmed that the BP-TFG bio-sensor could be used to detect the NSE/anti-NSE binding for multiple times.
3.5. Specificity of BP-TFG biosensor
The specificity of the biosensor was evaluated to detect the non-specific analytes such as IgG, PSA and PBS in comparison with the specific target NSE, all at a same concentration of 10 ng/mL. During the detection processes, the optical signal was monitored in real-time and recorded (Fig. S6). The comparison results are presented in Fig. 6c, showing the specificity evaluation with the initial reaction rate of the first 3-min and the maximum signal change over whole process. The initial rates for PBS, IgG and PSA were 0%, 6%, 3% that of NSE, while the maximum wavelength shifts for PBS, IgG and PSA were 0%, 25%,
19% that of NSE, respectively. This indicates that the proposed bio-sensor shows a strong preference for the specific affinity binding of NSE, demonstrating sufficient selectivity for NSE detection.
4. Conclusion
We proposed the first BP-fiber optic biosensor for ultrasensitive diagnosis of NSE cancer biomarkers. BP nanosheets were synthesized and deposited on fiber device to enhance the light-matter interaction where the unique optical modulation effects induced by BP were ex-perimentally observed. The PLL-biofunctionalized BP provided a re-markable analytical platform for affinity binding interface. The BP-TFG was implemented to detect NSE biomarkers demonstrating an ultrahigh sensitivity with the LOD of 1.0 pg/mL, which is 4 orders magnitude lower than NSE cut-off value of SCLC. The enhanced sensitivity is 100-fold higher than GO- or AuNPs-based biosensors. The capability of BP-fiber optic biosensor with ultrahigh sensitivity and specificity opens up the possibility for early diagnosis of cancer, tumor and diseases. In order to match the requirements of practical application, the perfor-mance of BP-TFG biosensor will be evaluated with the clinical samples in the future work.