广西科学  2016, Vol. 23 Issue (4): 378-380   PDF    
具有体液免疫反应的传染病模型稳定性分析
田海燕, 陈富, 康淑瑰     
山西大同大学数学与计算机科学学院, 山西大同 037009
摘要: 建立考虑潜伏感染细胞并具有体液免疫反应的传染病模型,讨论其解的非负性和有界性,得到确定模型动力学性态的基本再生数,再通过构造适当的Lyapunov泛函,并利用LaSalle不变原理证明模型无病平衡点的全局渐近稳定性.
关键词: 病毒感染     全局稳定性     免疫反应     李雅普诺夫函数    
Stability Analysis of an Infection Model with Humoral Immunity
TIAN Haiyan , CHEN Fu , KANG Shugui     
School of Mathematics and Computer Science, Datong University, Datong, Shanxi, 037009, China
Abstract: In this paper, we built a virus dynamics model with humoral immune response including latently infected cells, and then discussed the nonnegativity and boundedness of the solution.The basic reproduction number was obtained, which determined the dynamical behaviors of the infection model.By constructing suitable Lyapunov functions and applying LaSalle's invariance principle we proved that the infection-free equilibrium was globally asymptotically stable.
Key words: virus infection     global stability     immune response     Lyapunov function    
0 引言

近年来,越来越多的学者利用数学模型分析宿主细胞和病毒之间的相互作用,研究过的病毒有人类免疫缺陷病毒(HIV), 乙肝病毒(HBV),丙肝病毒(HCV), SEIR传染病毒等[1-7].为了描述易感染细胞,感染细胞以及乙肝病毒颗粒之间的关系,建立了如下基本的病毒动力学模型:

$ \left\{ {\begin{array}{*{20}{l}} {\frac{{{\rm{d}}x}}{{{\rm{d}}t}} = \lambda - ax - \beta xz,}\\ {\frac{{{\rm{d}}y}}{{{\rm{d}}t}} = \beta xz - by,}\\ {\frac{{{\rm{d}}v}}{{{\rm{d}}t}} = cy - dv,} \end{array}} \right. $ (1)

然而要建立更精确的病毒感染数学模型,必须考虑免疫反应.免疫系统中对病毒感染有影响的两个主要反应是细胞免疫和体液免疫.细胞免疫中细胞毒性T淋巴细胞在病毒防御中起着关键的作用,因为细胞毒性T淋巴细胞可以攻击并杀死被感染细胞.而体液免疫是基于B细胞产生的抗体攻击并杀死被感染细胞.在一些病毒感染中,比如疟疾,细胞免疫比体液免疫的效果差[8-10].Muras等[9]已经建立了体液免疫的基本动力学模型:

$ \left\{ {\begin{array}{*{20}{l}} {\frac{{{\rm{d}}x}}{{{\rm{d}}t}} = \lambda - dx - \beta xv,}\\ {\frac{{{\rm{d}}y}}{{{\rm{d}}t}} = \beta xv - by,}\\ {\frac{{{\rm{d}}v}}{{{\rm{d}}t}} = ky - uv - rzv,}\\ {\frac{{{\rm{d}}z}}{{{\rm{d}}t}} = gzv - \mu z.} \end{array}} \right. $ (2)

其中x, y, v分别表示易感染细胞,受感染细胞以及游离病毒颗粒的数量,z表示B细胞的数量.未感染细胞以常速率λ产生,死亡速率dx,细胞被感染的速率为βxv,受感染的细胞死亡率为by.游离病毒颗粒从受感染的细胞中产生的速率为ky,死亡率为uv,同时被抗体作用移除体内的速率为rzv.B细胞被激活的速率为gzv,死亡率为μz.所有系数为正.

但是系统(2)并未考虑潜伏的感染细胞(细胞的感染率以双线性函数βxv的形式给出).2009年,Gang Huang等[11]建立具有Bsddington-DeAngelis型功能性反应函数的病毒动力学模型,用更具一般形式的功能性反应函数$ \frac{{\beta xv}}{{1 + \gamma x + \alpha v}} $代替模型(2)中双线性函数βxv.从生物学的角度来看, 该模型更能描述病毒与正常细胞之间的动态演化作用.然而文献[10]未考虑潜伏的感染细胞和体液免疫反应.文献[12]在文献[11]的基础上,建立考虑潜伏感染细胞并具有Bsddington-DeAngelis型功能性反应函数的病毒动力学模型,但未考虑体液免疫反应。本文在文献[8-12]的启发下,建立具有Bsddington-DeAngelis型功能性反应函数传染病模型,既考虑体液免疫反应又考虑潜伏的感染细胞,模型如下:

$ \left\{ {\begin{array}{*{20}{l}} {\frac{{{\rm{d}}x}}{{{\rm{d}}t}} = \lambda - dx - \frac{{\beta xv}}{{1 + \gamma x + \alpha v}},}\\ {\frac{{{\rm{d}}x}}{{{\rm{d}}t}} = \frac{{\left( {1 - q} \right)\beta xv}}{{1 + \gamma x + \alpha v}} - \left( {e + \delta } \right)w,}\\ {\frac{{{\rm{d}}y}}{{{\rm{d}}t}} = \frac{{q\beta xv}}{{1 + \gamma x + \alpha v}} + \delta w - by,}\\ {\frac{{{\rm{d}}v}}{{{\rm{d}}t}} = ky - uv - rzv,}\\ {\frac{{{\rm{d}}z}}{{{\rm{d}}t}} = gzv - \mu z.} \end{array}} \right. $ (3)

其中γα是正常数,wy分别表示潜伏的感染细胞和活动性感染细胞的浓度.潜伏感染细胞死亡的速率ew,转化为活动性感染细胞的速率δw.1-qq(0 < q < 1)分别表示受病毒感染的细胞中变成潜伏细胞或活动细胞的概率,其他变量与系统(2)有相同的含义.

1 模型基本性质

系统(3)的所有解都非负有界.解的非负性从生物学意义上讲显然成立,只需证明其有界性.

定理1系统(3)的所有解x(t), w(t), y(t), v(t), z(t)都是非负且有界的,即存在Mi > 0, i=1, 2, 3使得0≤x(t), w(t), y(t)≤M1,0≤v(t)≤M2,0≤z(t)≤M3成立.

证明X1(t)=x(t)+w(t)+y(t),则有

$ \frac{{{\rm{d}}{X_1}\left( t \right)}}{{{\rm{d}}t}} = \lambda-dx\left( t \right)-ew\left( t \right)by\left( t \right) < \lambda-\\ {s_1}{X_1}\left( t \right). $

这里s1=min {d, b, e}.因此,对充分大的t,有X1(t) < M1+ε(ε为任意小的正数),其中$ {M_1} = \frac{\lambda }{{{s_1}}} $.所以0≤x(t), w(t), y(t)≤M1.

另一方面,设$ {X_2} = v\left( t \right) + \frac{r}{g}z\left( t \right), $则有

$ \frac{{{\rm{d}}{X_2}\left( t \right)}}{{{\rm{d}}t}} = ky\left( t \right)-uv\left( t \right)-\frac{{r\mu }}{g}z\left( t \right) \le k{M_1}-\\ {s_2}\left( {v\left( t \right) + \frac{r}{g}z\left( t \right)} \right) = k{M_1} -{s_2}{X_2}\left( t \right). $

这里s2=min {u, μ}.因此,对充分大的t,有X2(t) < M2+ε(ε为任意小的正数),其中$ {M_2} = \frac{{k{M_1}}}{{{s_2}}} $.所以0≤v(t)≤M2,0≤z(t)≤M3,其中$ {M_3} = \frac{{g{M_2}}}{r} $.

2 模型无病平衡点的稳定性

为方便,引入函数H:(0, ∞)→[0, ∞),定义为H(s)=s-1-lns.基本再生数R0定义为$ {R_0} = \frac{{k\beta {x_0}\left( {eq + \delta } \right)}}{{bu\left( {e + \delta } \right)\left( {1 + \gamma {x_0}} \right)}}, $其中$ {x_0} = \frac{\lambda }{d}. $

定理2R0≤1时,系统(3)的无病平衡点$ {E_0} = \left( {\frac{\lambda }{d}, 0, 0, 0, 0} \right) $是全局渐近稳定的.

证明构造Lyapunov函数:

$ \begin{array}{*{20}{l}} {\;\;\;\;\;\;\;{W_0} = \frac{{{x_0}}}{{1 + \gamma {x_0}}}H\left( {\frac{x}{{{x_0}}}} \right) + \frac{\delta }{{eq + \delta }}w + \frac{{e + \delta }}{{eq + \delta }}y}\\ {\frac{{b\left( {e + \delta } \right)}}{{k\left( {eq + \delta } \right)}}v + \frac{{br\left( {e + \delta } \right)}}{{kg\left( {eq + \delta } \right)}}z.} \end{array} $

结合系统(3)有

$ \begin{array}{*{20}{l}} {\frac{{{\rm{d}}{W_0}}}{{{\rm{d}}t}} = \frac{1}{{1 + \gamma {x_0}}}\left( {1 - \frac{{{x_0}}}{x}} \right)\left( {\lambda - dx - } \right.}\\ {\left. {\frac{{\beta xv}}{{1 + \gamma x + \alpha v}}} \right) + \frac{\delta }{{eq + \delta }}\left( {\frac{{\left( {1 - q} \right)\beta xv}}{{1 + \gamma x + \alpha v}} - \left( {e + } \right.} \right.}\\ {\left. {\left. \delta \right)w} \right) + \frac{{e + \delta }}{{eq + \delta }}\left( {\frac{{q\beta xv}}{{1 + \gamma x + \alpha v}} + \delta w - by} \right) + }\\ {\frac{{b\left( {e + \delta } \right)}}{{k\left( {eq + \delta } \right)}}\left( {ky - uv - rzv} \right) + \frac{{br\left( {e + \delta } \right)}}{{kg\left( {eq + \delta } \right)}}\left( {gzv - } \right.}\\ {\left. {\mu z} \right) = - d\frac{{{{\left( {x - {x_0}} \right)}^2}}}{{x\left( {1 + \gamma {x_0}} \right)}} - \frac{{\beta xv}}{{\left( {1 + \gamma {x_0}} \right)\left( {1 + \gamma x + \alpha v} \right)}} + }\\ {\frac{{\beta {x_0}v}}{{\left( {1 + \gamma {x_0}} \right)\left( {1 + \gamma x + \alpha v} \right)}} + \frac{{\beta xv}}{{1 + \gamma x + \alpha v}} - }\\ {\frac{{bu\left( {e + \delta } \right)}}{{k\left( {eq + \delta } \right)}}v - \frac{{br\mu \left( {e + \delta } \right)}}{{kg\left( {eq + \delta } \right)}}z = - d\frac{{{{\left( {x - {x_0}} \right)}^2}}}{{x\left( {1 + \gamma {x_0}} \right)}} + }\\ {\frac{{\beta {x_0}\left( {1 + \gamma x} \right)}}{{\left( {1 + \gamma {x_0}} \right)\left( {1 + \gamma x + \alpha v} \right)}} - \frac{{bu\left( {e + \delta } \right)}}{{k\left( {eq + \delta } \right)}}v - }\\ {\frac{{br\mu \left( {e + \delta } \right)}}{{kg\left( {eq + \delta } \right)}}z = - d\frac{{{{\left( {x - {x_0}} \right)}^2}}}{{x\left( {1 + \gamma {x_0}} \right)}} - \frac{{bu\left( {e + \delta } \right)}}{{k\left( {eq + \delta } \right)}} \bullet }\\ {\frac{{\alpha {v^2}{R_0}}}{{1 + \gamma x + \alpha v}} + \frac{{bu\left( {e + \delta } \right)}}{{k\left( {eq + \delta } \right)}}\left( {{R_0} - 1} \right)v - \frac{{br\mu \left( {e + \delta } \right)}}{{kg\left( {eq + \delta } \right)}}z.} \end{array} $ (4)

因此,当R0≤1时,对任意的x, v, z>0有$ \frac{{{\rm{d}}{W_0}}}{{{\rm{d}}t}} \le 0 $成立.又易知系统(3)的解收敛于一个集合Γ,其中Γ$ \left\{ {\frac{{{\rm{d}}{W_0}}}{{{\rm{d}}t}} \le 0} \right\} $的最大不变子集.同时由(4)式知,$ {\frac{{{\rm{d}}{W_0}}}{{{\rm{d}}t}} = 0} $当且仅当x=x0, v=0, z=0成立.注意到,对Γ中的任何元素,如果v=z=0,就有$ {\frac{{{\rm{d}}{W_0}}}{{{\rm{d}}t}} = 0} $.又由系统(3)的第4个方程知,$ 0 = \frac{{{\rm{d}}v}}{{{\rm{d}}t}} = ky $,因此y=0.此外,由系统(3)的第3个方程知w=0.因此,$ {\frac{{{\rm{d}}{W_0}}}{{{\rm{d}}t}} = 0} $当且仅当x=x0, y=0, v=0, z=0成立.故由时滞微分方程的LaSalle不变原理知,无病平衡点$ {E_0} = \left( {\frac{\lambda }{d}, 0, 0, 0, 0} \right) $)是全局渐近稳定的.

3 结论

本文讨论了一类考虑潜伏感染细胞的具有体液免疫反应的传染病模型,该模型描述了未受感染的靶细胞,潜伏感染的细胞,活动性感染的细胞,自由的病毒颗粒与B细胞之间的相互作用.利用Lyapunov函数和LaSalle不变原理,证明了当基本再生数R0≤1时,模型无病平衡点全局渐近稳定.

参考文献
[1]
谭光兴, 陈少白, 毛宗源. 一类时滞免疫反应系统稳定性分析[J]. 广西科学, 2006, 13(1): 19-22.
TAN G X, CHEN S B, MAO Z Y. Stability analyze in a class of time-delay immune response system[J]. Guangxi Sciences, 2006, 13(1): 19-22.
[2]
GOURLEY S A, KUANG Y, NAGY J D. Dynamics of a delay differential equation model of hepatitis B virus infection[J]. Journal of Biological Dynamics, 2008, 2(2): 140-153. DOI:10.1080/17513750701769873
[3]
韩晓丽, 靳祯. 具有免疫反应的乙肝病毒时滞动力学模型[J]. 中北大学学报:自然科学版, 2011, 32(1): 35-38.
HAN X L, JIN Z. A dynamic model of hepatitis B virus with delayed immune response[J]. Journal of North University of China:Natural Science Edition, 2011, 32(1): 35-38.
[4]
NOWAK M A, BONHOEFFER S, HILL A M, et al. Viral dynamics in hepatitis B virus infection[J]. Proceedings of the National academy of Sciences of the United States of America, 1996, 93(9): 4398-4402. DOI:10.1073/pnas.93.9.4398
[5]
庄科俊. 一类时滞乙肝病毒模型的稳定性分析[J]. 中北大学学报:自然科学版, 2015, 36(2): 122-125.
ZHUANG K J. Stability analysis for a hepatitis B virus model with time delay[J]. Journal of North University of China:Natural Science Edition, 2015, 36(2): 122-125.
[6]
KOROBEINIKOV A. Global properties of basic virus dynamics models[J]. Bulletin of Mathematical Biology, 2014, 66(4): 879-883.
[7]
徐翠翠. 一类SEIR传染病模型周期解的存在性[J]. 广西科学, 2011, 18(1): 17-21.
[8]
XU C C. Periodic solution of a SEIR epidemic model[J]. Guangxi Sciences, 2011, 18(1): 17-21.
[9]
OBAID M A, ELAIW A M. Stability of virus infection models with antibodies and chronically infected cells[J]. Abstract and Applied Analysis, 2014, 2014: 650371.
[10]
MURASE A, SASAKI T, KAJIWARA T. Stability analysis of pathogen-immune interaction dynamics[J]. Journal of Mathematical Biology, 2005, 51(3): 247-267. DOI:10.1007/s00285-005-0321-y
[11]
WANG S F, ZOU D Y. Global stability of in-host viral models with humoral immunity and intracellular delays[J]. Applied Mathematical Modelling, 2012, 36(3): 1313-1322. DOI:10.1016/j.apm.2011.07.086
[12]
HUANG G, MA W B, TAKEUCHI Y. Global properties for virus dynamics model with Beddington-DeAngelis functional response[J]. Applied Mathematics Letters, 2009, 22(11): 1690-1693. DOI:10.1016/j.aml.2009.06.004
[13]
ELAIW A M, AZOZ S A. Global properties of a class of HIV infection models with Beddington-DeAngelis functional response[J]. Mathematical Methods in the Applied Sciences, 2013, 36(4): 383-394. DOI:10.1002/mma.v36.4